Q: Recent research has defined mild cognitive
impairment as a transitional state between the
cognitive changes of normal aging and Alzheimer
disease (AD) and other dementing illnesses. What
criteria are used to differentiate mild cognitive
impairment from more innocuous syndromes, such as
benign senescent forgetfulness? Are patients with mild
cognitive impairment considered to have incipient
A: Benign senescent forgetfulness (age-associated memory
impairment)-which results from the slowing of
neural processes with age-is thought to be an extreme of
normal aging. It does not significantly interfere with activities
of daily living and is not a precursor of pathologic aging.
As we age, we all notice an increase in "senior moments"-
such as forgetting someone's name or why we went into
a particular room. Benign senescent forgetfulness usually
affects short-term memory rather than long-term or remote
memory. Impaired learning ability manifests only in the setting
of timed testing. Persons with benign forgetfulness are
usually more concerned about this than family members. If
a patient complains of being forgetful, he or she is less likely
to have a dementia than if a family member reports it and
the patient denies it.
Mild cognitive impairment can be characterized as
memory impairment without dementia (Table 1); that is,
patients experience memory loss, but other higher cortical
functions remain intact and none of the other features that
suggest dementia (Table 2) are present. Patients may manifest
memory impairment that is more extensive than would
be expected for their age or have some minor word-finding
difficulties. Mild cognitive impairment can be documented
on objective tests of memory.1 However, general cognitive
function is preserved and activities of daily living, social relationships,
and occupational responsibilities are not affected.
Therefore, patients with mild cognitive impairment do
not meet criteria for clinically probable AD. It is not yet
known whether mild cognitive impairment represents the
earliest stages of AD or is a distinct clinical entity.
Persons with benign senescent forgetfulness have a 1%
to 2% per-year risk of progressing to dementia-the same
as that for all persons 65 years or older. However, clinically
probable AD develops in up to 25% of persons with mild
cognitive impairment per year.2,3Q: What is the best way to evaluate a patient who
presents with memory impairment-or whose
family is concerned about this?
A: The first step is to rule out dementia, a condition that
always involves a decline from a previous level of intellectual
functioning.4 Dementia is a clinical diagnosis
based on a history of cognitive decline and the absence of
other conditions, such as delirium or depression. Screening
tests-even if they are adjusted for educational level-are
not diagnostic of dementia.
The diagnosis of dementia is made following documentation
of memory impairment (usually short-term
memory), which typically manifests first, and at least one
other cognitive disturbance, such as apraxia, agnosia, aphasia,
or impaired executive function. Each of the cognitive
deficits causes significant social or occupational dysfunction
and affects activities of daily living.
Early personality changes are more
characteristic of AD, whereas early
emotional lability is usually associated
with multi-infarct dementias. Frontal
lobe dementias typically present with
depressive symptoms or excessive behaviors.
Poor judgment can occur with
all types of dementia; it may result
from a combination of cognitive deficits
or the patient's attempts to hide
Q: How valuable are the screening tests used to
evaluate patients with cognitive impairment?
A: The Mini-Mental State Examination (MMSE) is probably
the most commonly used instrument to screen
for cognitive impairments. The MMSE (which also exists in
shorter versions) is likely to be particularly useful for detecting
early dementia, before the onset of functional and
behavioral disturbances. However, screening instruments
are not sufficient by themselves to make a diagnosis of dementia,
in part because performance on such tests is a function
of many factors, including education, age, cultural background,
and comorbid conditions. Formal neuropsychological
testing may be required to support
a clinical diagnosis of dementia or mild
Although a screening test cannot
alone make a diagnosis of dementia,
results can guide the investigation of
other diagnostic possibilities. If a patient
scores between 24 and 26 on the
MMSE (the normal range is above 24,
and 30 is the highest score), he may
still meet the criteria for dementia. The
MMSE primarily tests different aspects
of orientation, language, and memory; the test of constructional
abilities counts for only 1 of 30 points. The Clock
Drawing Test further tests these abilities, as well as more
complex higher cognitive functions, and is a useful adjunct
to the MMSE (Figure).
Persons with benign senescent forgetfulness should
score in the normal range on screening cognitive tests (except
for timed testing exercises), because their ability to
learn new information is not impaired. Persons with mild
cognitive impairment might get only 1 or 2 of 3 objects cor-
rect on a 3-object recall but should be able to perform well
on other areas of the MMSE.
Screening tests can be administered at different points
over the course of a dementing illness; thus, they can help
track disease progression. Since the rate of cognitive decline
varies greatly, these tests enable you to document changes
objectively. A concomitant problem-such as an adverse effect
of a medication-may escalate the rate of decline. When
this effect is identified and the drug is withdrawn, the related
cognitive changes may reverse themselves.
Q: Does current research
support an intervention-
such as pharmacotherapy-early in
the course of mild cognitive
A: Some researchers advocate
starting therapy with a cognitiveenhancing
agent, such as a cholinesterase
inhibitor, for patients with mild
cognitive impairment. The reasoning is that since there is a
high conversion rate from mild cognitive impairment to
AD, it is likely that many patients with mild cognitive impairment
have the underlying neuropathology of AD, even
though they do not yet meet clinical diagnostic criteria.5
This question remains controversial, however. Some
research suggests that cholinesterase inhibitors may slow
the progression of the underlying pathology while providing
symptomatic management for patients with dementia.
However, there is no evidence that mild cognitive impairment
can be successfully treated.1 Thus, I do not recommend
that cholinesterase inhibitors be used until there is an
actual diagnosis of dementia. In the case of a patient with
mild cognitive impairment, I first rule
out medication side effects and depression,
document the cognitive
changes, administer the appropriate
screening tests, and have the patient
return in 6 months to assess the progression
of the documented changes.
Q: There is some debate about
the value of neuroimaging in
patients with mild cognitive
impairment. What is its role?
A: In the absence of neurologic
findings, initial neuroimaging
studies in a patient with mild cognitive
impairment are of limited value. Significant
hippocampal and entorhinal
cortex atrophy has been found in patients with mild cognitive
impairment; such changes are not found in healthy persons.
However, these changes do not have a high predictive
value for individual patients.6
Consider CT or MRI for a patient with atypical features
of dementia or for a patient with dementia who is experiencing
a period of cognitive decline that is more rapid than
would be expected. Atypical features or rapid decline may
suggest another pathology-such as normal-pressure hydrocephalus,
which is a treatable component of dementia.
Q: What common causes of
memory impairment are
most likely to mimic dementia?
How can I efficiently rule them out?
A: Drug reactions, depression, and
delirium must be ruled out before
a diagnosis of dementia can be
made. More than half of treatable dementias
are attributable to either depression
or an adverse effect of medication. Unfortunately,
it is very tempting to prescribe a cognitive-enhancing drug
if a patient complains of memory impairment. However, it
is crucial to be absolutely certain of the diagnosis to ensure
appropriate management. Even in a patient with diagnosed
dementia, it is worth reviewing his medications and discontinuing
those with adverse cognitive effects, to optimize
the benefits of a cognitive-enhancing agent.
Drug reactions. If your patient presents with a memory
complaint, be sure that the history includes a complete
list of medications he is taking, including over-the-counter
remedies, complementary and alternative agents, and any
"borrowed"drugs. The medications most closely associat-
ed with memory impairment are benzodiazepines and narcotic
analgesics. Agents that have anticholinergic side effects-
including some harmless-seeming ones, such as allergy
medications and H2 blockers like ranitidine-can also
impair memory temporarily.
In most cases, drug-related memory impairment develops
gradually, as metabolites build up. This is particularly
the case with benzodiazepines, such as lorazepam and alprazolam,
which are both fat-soluble; the changes can be
fairly subtle and the patient may not be aware of them. If
you suspect a drug reaction, taper and/or stop the suspected
agent immediately. Memory function generally returns
to normal within a week or so of drug discontinuation, and
the change is very obvious both to the patient and to his
family members or caregivers.
Depression. Memory complaints
can sometimes confuse the clinical picture
in a patient with depression. Depressed
elderly persons usually report
low energy, loss of appetite, and sleep
problems. They feel worst when they
get up in the morning. Memory complaints
in such patients are often subjective
and reflect an impaired ability
to concentrate, but may result in "I
don't know" answers on objective
tests. Even if you suspect coexisting
dementia (documented depression in
the early stages of dementia is common),
it is wise to treat depression first
in these persons.
Delirium. Dementia is a risk factor
for delirium. In general, delirium
has a relatively sudden onset (over days to a couple of
weeks in outpatients), is usually reversible, and is often associated
with medication use or withdrawal or with an acute
illness. Dementia is characterized by insidious onset and
gradual decline over a long period (typically more than 6
months); it is not usually associated with a pharmacologic
agent or an acute illness.
If a patient's family reports that he had an acute episode
of confusion during hospitalization and relates this to the
onset of cognitive decline, it may be difficult to be certain of
the underlying diagnosis. The first step is to manage any factors
that might cause delirium; the next is to treat suspected
depression before moving on to confirm a diagnosis of dementia
and initiating cognitive-enhancing therapy.
Q: Can you recommend any guidelines for helping
a patient withdraw from a drug, such as a
benzodiazepine, that may be responsible for memory
A: It can, of course, be difficult to wean patients off certain
agents-particularly the benzodiazepines, such as
alprazolam. These agents can be quite addictive, and patients
can experience withdrawal symptoms, such as insomnia,
anxiety, or even seizures.7 The taper needs to be done quite
slowly to ensure compliance; the process may take as long
as 3 to 6 months for patients who have been using elevated
dosages. It is essential to have the patient's cooperation in
the withdrawal process. I have had good results by assigning
the patient the responsibility of determining the rate of
the taper, using goal setting as an incentive.
In some of my patients who have been taking a benzodiazepine
(particularly alprazolam), I taper the dosage
down to lower levels and then substitute oxazepam. Because
this agent is water-soluble rather
than lipid-soluble, its onset is slower,
its duration of action is predictable, it
has no active metabolites, and it is less
addictive. Temazepam is an alternative
water-soluble benzodiazepine that is a
particularly effective replacement for
nighttime sedatives for insomnia.
Q: We've all had patients who
insist that without their
benzodiazepines they have
unacceptable levels of anxiety or
insomnia. How do you address
A: I explain to patients that although
they may have been
taking a particular drug without any
apparent side effects for a long time, aging of the brain diminishes
the ability to tolerate the medication as before.
I tell them that these drugs can cause cognitive impairment
and also increase the risk of depression or delirium.
I emphasize that rather than concentrating on the loss of
the insomnia medication or anxiolytic, they should view
the process as an investment in the recovery of their
Q: What is a good substitute if patients require a
sleep aid during the period in which a
benzodiazepine is being withdrawn?
A: I initially prescribe a nightly dose of trazodone, a mild
antidepressant with very good sedative properties. I
start with 25 mg for the first 4 days and then titrate up to 50
or 100 mg if necessary, over a period of a few weeks. This
strategy ensures efficacy without daytime somnolence. Ultimately,
the goal is to have the patient discontinue all sleep-
ing medications, but it's a lot easier-and probably safer-
to wean a patient off trazodone than some of the other
agents; for example, benzodiazepine withdrawal symptoms
do not occur with trazodone.
Some patients find that alcohol helps induce sleep, but
alcohol disrupts sleep architecture and may also contribute
to cognitive impairment. Alcohol use is best discouraged in
patients with this condition.
Q: Is there evidence that supports (or refutes) the
use of ginkgo biloba to prevent or treat
memory disorders? Some of my patients start taking it
when they notice that they are becoming forgetful.
A: A recent randomized, double-blind, placebo-controlled
6-week trial of a ginkgo product found that it did not
improve memory or related cognitive
function in cognitively normal adults
older than 60 years.8 A recent metaanalysis
by Birks and colleagues9 found
that there was some evidence of improvement
in cognition and function associated
with ginkgo among patients
with cognitive impairment. However,
these authors noted that many of the
trials were small and methodologically
flawed and that publication bias could
not be ruled out. They also observed
that the results of more recent trials
have been ambiguous.
Certainly one concern about using ginkgo biloba is the
lack of batch-to-batch consistency. Some of my patients who
have taken ginkgo seem to derive real benefit from it, although
the objective evidence of benefit is anecdotal at best.
The powerful placebo effect of this agent (as high as 50% in
some trials) has been demonstrated repeatedly and may result
from changes at the neurotransmitter level.
Some research suggests that a change in a patient's
environment or the way he feels about himself can be as effective
as medication. Such psychological factors have very
powerful effects that we don't completely understand.
1. Petersen RC, Doody R, Kurz A, et al. Current concepts in mild cognitive impairment.
Arch Neurol. 2001;58:1985-1992.
2. Petersen RC, Smith GE, Waring SC, et al. Mild cognitive impairment: clinical
characterization and outcome. Arch Neurol. 1999;56:303-308.
3. Petersen RC, Stevens JC, Ganguli M, et al. Practice parameter: early detection
of dementia: mild cognitive impairment (an evidence-based review).
4. Beers MH, Berkow R, eds. The Merck Manual of Geriatrics. 3rd ed. Whitehouse
Station, NJ: Merck & Co; 2000.
5. Sramek JJ, Veroff AE, Cutler NR. The status of ongoing trials for mild cognitive
impairment. Expert Opin Investig Drugs. 2002;10:741-752.
6.Wolf H, Jelic V, Gertz HG, et al. A critical discussion of the role of neuroimaging
in mild cognitive impairment. Acta Neurol Scand Suppl. 2003;179:52-76.
7. Juergens S. Alprazolam and diazepam: addiction potential. J Subst Abuse Treat.
8. Solomon PR, Adams F, Silver A, et al. Ginkgo for memory enhancement: a
randomized controlled trial. JAMA. 2002;288:835-840.
9. Birks J, Grimley EV, Van Dongen M. Ginkgo biloba for cognitive impairment
and dementia. Cochrane Database Syst Rev. 2002;(4):CD003120.
10. Schneiderman H. Strong will strikingly compensating for visual field cut and
hemineglect on Albert's test after nondominant parietal stroke. Consultant. 2003;
FOR MORE INFORMATION:
- Chetelat G, Baron JC. Early diagnosis of Alzheimer's disease: contribution of
structural neuroimaging. Neuroimage. 2003;18:525-541.
- De Smet PA. Herbal remedies. N Engl J Med. 2002;347:2046-2056.
- Hanninen T, Hallikainen M, Koivisto K, et al. A follow-up study of age-associated
memory impairment: neuropsychological predictors of dementia. J Am Geriatr
- Morris MC, Evans DA, Bienias JL, et al. Vitamin E and cognitive decline in
older persons. Arch Neurol. 2002;59:1125-1132.
- Thal LJ, Thomas RG, Mulnard R, et al. Estrogen levels do not correlate with
improvement in cognition. Arch Neurol. 2003;60:209-212.