A 74-Year-Old Man With Memory Loss and Neuropathy Who Enjoys Alcoholic Beverages

DR DELBANCO: Mr E is a 74-year-old man who enjoys drinking alcoholic beverages and is worried by memory loss and neuropathic symptoms. A retired attorney, he has Medicare and commercial insurance coverage. He is single and separated from his family. Mr E has not smoked and exercises regularly on a bike and a treadmill. He used to play tennis but finds he cannot really do so anymore. There is no family history of alcohol abuse or neuromuscular disease.

About 20 years ago, Mr E noticed that his hands were thinning and were not working well. For the past 10 years, he has found himself unable to open jars, and his hands look misshapen, with claw deformities of his last 2 digits bilaterally. He also developed hammer toes and “strange feelings” in the soles of his feet when walking.

Examination by a neurologist in 1998 revealed considerable atrophy of the intrinsic hand muscles, and Mr E was also thought to have bilateral ulnar entrapment. In addition, the consultant noted some thinning of the legs below the knee. Ulnar release surgery stabilized some of the sensations of weakness and paresthesias in his hands but did not return his function to normal. After further consultation, along with evidence from electromyograms, his physicians believed he had a slowly progressive, probably hereditary, demyelinating neuropathy that likely represented one of the Charcot-Marie-Tooth variants.

Mr E has enjoyed a “couple of heavy drinks daily” for much of his life and as much as “3 to 4 vodkas a day, sometimes more.” He ceased working abruptly about 10 years ago and found himself drinking more. In retrospect, he believes this increase in consumption resulted primarily from the void he experienced after leaving work and from having little to do in that time. Shortly thereafter, he found himself somewhat “down in the dumps” and began to drink quite heavily. He drank “real liquor, not wine, not beer.”

During the past 10 years, he has noted some memory loss. He has increasing trouble remembering names. There are times he cannot find words, and he believes he does not think as quickly as he used to. He denies blackouts or hallucinations. He has also developed hypertension, which has been difficult to manage at times. He has been generally healthy other than requiring 3 herniorrhaphies. Currently, he has cut down on his drinking but still has 1 or 2 drinks a day. He takes verapamil, hydrochlorothiazide, and lisinopril for hypertension and ranitidine for gastroesophageal reflux disease.

On a recent visit to a neurologist, his affect was full, although he spoke rather slowly. There was no further comment on his mental status. He looked well. His blood pressure was 130/80 mm Hg and has been normal at home as well. His pulse was 74/min and regular, and he was not tachypneic. He had moderate hearing loss, and one of his tympanic membranes was scarred. The general physical examination was unremarkable; he did not have findings suggestive of liver disease or skin changes associated with alcohol abuse.

On recent neurologic evaluations, the cranial nerves appeared normal and there was no tremor. Finger-to-nose repetitions were not remarkable. There was marked atrophy of the intrinsic muscles in his hands and claw deformities of the fourth and fifth fingers bilaterally (Figure; see video at http://jama.ama-assn.org/cgi/content/full/299/9/1046/DC1). He had moderate muscular atrophy and diminished perception of vibration, pin prick, and temperature below his knees. Deep tendon reflexes were absent in the lower extremities. Fasciculations were noted once, but not persistently, and his toes were down-going. Mr E was unable to maintain tandem gait, and his walking was wide-based (Figure; see video at http://jama.ama-assn.org/cgi/content/full/299/9/1046/DC1).

Laboratory evaluation revealed a normal complete blood cell count. His platelet count and international normalized ratio have been normal. In the past, his mean corpuscular volume (MCV) was elevated at times. Vitamin B₁₂ levels have been normal recently, although when first evaluated years ago they were borderline low, and for a while he had injections of vitamin B₁₂. Liver function test results have been normal. He has had persistently high creatine phosphokinase levels. Thyrotropin, immunoglobulins, and folate levels have been normal. His blood test for human immunodeficiency virus is negative.

I worked with a major international company, which had a very strict policy about alcohol. And there was no alcohol on the premises. If you went out and had a dinner in a group, it was a rare case where you would actually have wine or alcohol at the cost of the company. I didn't dive into the bottle the day after I quit. But the issue really was that I ended work fairly abruptly and I had not really focused on what I was going to do after that. So there was a gap in my time and in what I had to think about. And I think that's where I got started.

I started out drinking more in the evening. And after a while I started to have a drink at noontime. I really have problems in telling you how much I drink. I have never been dead drunk. I have never fallen. I occasionally fall asleep after dinner when watching TV, or something like that. And the next morning I don't have a headache or feel bad, but I feel logier than I would if I had not drunk a lot. People have talked to me about it. Mostly it's been my daughter, who is a grown-up woman and thinks that I should cut down on it.

In some ways, perhaps, my motion is not the same, but it did not happen overnight. So this is one of these things that, if it happens, is slow and long-term. I and my doctor talk about feet and feeling in feet; each time I go see him he takes a pin and pokes around in the foot to see whether I’m losing some reflexes or feeling. I occasionally feel better when there is a handrail or something like that. I can't tell exactly why that is, but I think at this stage it is more a habit than a need or requirement. Or maybe I’m fooling myself and it is the beginning of a requirement. I feel definitely that I’ve lost a lot of strength in my hands. And there are things that I could do in the past, like pushing an on-off switch. I cannot do that unless I work very hard at it, or use 2 fingers, or use another finger, or something like that.

The brain may not be quite as fast and not as sharp. I also have very clearly, I think, developed a loss of memory. This may be a silly example, but I’m running around the house and I want to do 3 things. And I start. And I do the 2, and I forget what the third one was. So I’ve got to go back and recycle, something like that. Analyzing something, as a lawyer is supposed to do, every so often when I’m reading something about a case, I will miss points that I would probably have found, would have come up with almost automatically in my mind. Whether that's alcohol or something else, or just age, I don't know. It's hard to say because it is very irregular.

How does alcohol damage the nervous system and what is known about the prevalence of such damage? What are the principal causes and diagnostic considerations in evaluating nervous system abnormalities in a 74-year-old man who is an active drinker? How does one decide if the symptoms or signs are related to alcohol? As far as cognition is concerned, does alcohol have a “safe dose” threshold? What do you recommend for Mr E?

DR BRUST: For more than a dozen years, Mr E has had a slowly progressive sensorimotor polyneuropathy, demyelinating in type but otherwise of unclear etiology. A borderline serum cobalamin level several years ago was treated with vitamin B₁₂ replacement. In the past few years, he has had difficulty with recent memory—forgetting what he was going to say, hesitating with word-finding, and having to keep lists. Following his retirement more than a decade ago, he gradually increased his alcohol intake to a degree he recognizes as excessive. Among diagnostic uncertainties is whether alcohol is contributing to either his polyneuropathy or his cognitive difficulty.

Alcohol damages the nervous system in myriad ways, and while the prevalence of such complications is difficult to estimate, they are likely underrecognized (Box 1). Thiamine and, probably, additional vitamin deficiencies cause or contribute heavily to Wernicke-Korsakoff syndrome, cerebellar degeneration, polyneuropathy, and amblyopia. Quiz Ref IDPellagra, which causes cognitive impairment and psychiatric symptoms, is the result of nicotinic acid deficiency. Relevant to Mr E, alcoholic/nutritional polyneuropathy is a primarily axonal process, with only modestly reduced nerve conduction velocities.³

An early symptom of his polyneuropathy was weakness, prominent enough to support a tentative diagnosis of hereditary sensorimotor neuropathy of Charcot-Marie-Tooth type.¹⁷ Polyneuropathy in persons with alcoholism, in contrast, produces early sensory loss, often painful, with weakness usually a late feature. Mr E's neuropathic symptoms, moreover, began years before he began drinking excessively, and his history does not point to nutritional deficiency. On the other hand, direct toxicity probably contributes to the polyneuropathy of many heavy drinkers, and it is theoretically possible that it could be aggravating Mr E's neuropathic symptoms.

On examination, Mr E demonstrated a wide-based gait, which could be the result of polyneuropathy, especially if proprioception is impaired. Like weakness, proprioceptive loss tends to be a late feature of polyneuropathy in alcoholism. Gait and truncal ataxia in such patients is more often attributable to degeneration of the cerebellar vermis (which, like polyneuropathy, is usually the result of both nutritional deficiency and direct neurotoxicity).¹⁸

Indirectly, alcohol can damage the nervous system by causing or increasing the risk of hepatic failure, hypoglycemia, alcoholic ketoacidosis, infection, neoplasm, trauma, and stroke (Box 1). Each of these disorders can cause cognitive impairment, but there is nothing in Mr E's history to suggest them.

Less understood are alcoholic myopathy, Marchiafava-Bignami disease (a poorly understood leukoencephalopathy affecting especially the corpus callosum), nonnutritional alcoholic dementia, and fetal alcohol effects (Box 1). Direct neurotoxicity appears to play a major role in these disorders.

The principal diagnostic considerations in a 74-year-old heavy drinker with impaired memory are mild cognitive impairment of aging (MCI), Alzheimer disease, other dementing illnesses, depression, and effects of alcohol. How do clinicians tell them apart?

Quiz Ref IDMild cognitive impairment refers to impairment in cognitive function, typically memory, of a greater extent than would be expected for age but falling short of dementia, defined in the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition [DSM-IV]) as cognitive impairment involving more than memory and severe enough to interfere with occupational or social functioning.¹⁹ A degree of cognitive decline can be a part of normal aging, and Alzheimer disease begins insidiously, most often with subtle lapses of memory. Considerable investigation therefore has attempted to define the boundaries of MCI (ie, where normal aging ends and where Alzheimer dementia begins), but consensus regarding the specific criteria for MCI (and, therefore, its prevalence) has yet to be achieved.^{20 - 21}

By the time Alzheimer disease produces symptoms that meet criteria for dementia, deficits in more than 1 cognitive domain are present, including, variably, attention, language, calculation, spatial ability, executive function, and abstraction. On the other hand, some workers divide MCI into subgroups based on the presence of either impaired memory (amnestic) without other cognitive dysfunction, impaired memory plus other cognitive dysfunction, or other cognitive dysfunction without impaired memory and involving 1 or more nonamnestic cognitive domain.²² In 1 large series, nonamnestic MCI was as common as amnestic MCI.²³ Impairment in more than 1 cognitive domain thus cannot be used to distinguish MCI and early Alzheimer disease.

Depending on the criteria used to define MCI, cohort studies have reported that the rate of progression to dementia is between 6% and 25% per year, with most studies reporting between 10% and 15%.^{20 ,24 - 28} Conversion to dementia, most often Alzheimer disease, is thus nearly 80% within several years. In the Leipzig Longitudinal Study of the Aged (LEILA 75+), which assessed 980 community-dwelling individuals aged 75 years or older, a diagnosis of MCI was based on subjective report of cognitive impairment, preserved ability to perform activities of daily living, and absence of dementia according to DSM-IV criteria.²³ At baseline, 19.2% of the participants had MCI (4.5% amnestic in a single domain, 5.5% amnestic in multiple domains, 7.1% nonamnestic in a single domain, and 2.1% nonamnestic in multiple domains). Quiz Ref IDDuring an average 4.3 years of observation, 44% of individuals with a diagnosis of MCI developed dementia compared with 17% of those who were cognitively unimpaired. Those with amnestic MCI had a greater likelihood of progressing to dementia than those with nonamnestic types. In the Cardiovascular Health Study Cognition Study, amnestic MCI in individuals with associated disorders (psychiatric, neurological, or systemic) that could cause cognitive deficits was less likely to progress to dementia than in those without comorbid conditions (41% vs 70%).²⁹ Using variable criteria, epidemiological studies have found that 4% to 44% of individuals with MCI not only do not progress to dementia but revert to normal cognition.^{23 ,29 - 32}

Neuropsychological testing can identify the presence, type, and degree of cognitive impairment and can measure changes in cognition over time, but in the early stages of MCI it cannot predict whether symptoms will progress to dementia.²⁰ Investigators have described the usefulness of neuroimaging, electrophysiological studies, or biomarkers, but such technology is not widely available and their role clinically is not yet known (Box 2).^{33 - 38} At present, predicting the course of MCI in an individual is problematic.

Magnetic resonance imaging measuring hippocampal and entorhinal atrophy: During 1- to 9-year follow-up, patients with MCI and smaller cortex volumes were significantly more likely to progress to Alzheimer disease than patients with MCI and larger volumes or cognitively healthy controls with normal volumes (P < .001).³³

Positron emission tomography, using FDDNP, a radiotracer molecule that binds to amyloid and tau protein: Significantly greater binding occurred in patients with Alzheimer disease than in patients with MCI and in patients with MCI than in cognitively healthy controls (P < .001).³⁴

Magnetic encephalographic measurement of left parietal delta dipole density (LPD): MCI patients with high LPD were significantly more likely to develop Alzheimer disease within 2 years than those with low LPD (P = .002).³⁵

Event-related potential (ERP) assessment of a positive-negative ERP component during working memory (ERP-WM): MCI patients with no cognitive decline during a 1-year follow-up had visible ERP-WM at baseline. MCI patients with progressive cognitive decline did not (P value not given).³⁶

Measurement of plasma levels of amyloid β-protein 42: Levels in women with MCI were significantly higher than in cognitively healthy women (P < .05)³⁷ or men with MCI (P < .002).

Measurement of tau protein in cerebrospinal fluid: MCI patients with elevated tau protein levels at baseline were significantly more likely to have cognitive worsening during a 2-year follow-up than MCI patients with normal tau levels.³⁸

^aStudies have not been established to be useful in routine clinical care.

In patients with either MCI or frank dementia, diagnoses other than Alzheimer disease must be considered, especially those that are treatable. The most common causes of dementia in elderly individuals after Alzheimer disease are vascular dementia, Lewy body/Parkinson disease, alcoholism, and intoxication with drugs or medication. In a UK study of 1085 people aged 65 years or older, dementia was present in approximately 10%, of whom two-thirds were given a specific diagnosis. Thirty-one percent had probable Alzheimer disease, 22% had vascular dementia, and 11% had dementia with Lewy bodies.³⁹ A literature review reported a prevalence of dementia with Lewy bodies of 0% to 5% in the general population and of 0% to 35% among all dementia cases.⁴⁰

Multi-infarct dementia refers to cognitive impairment resulting from more than 1 discrete large-vessel stroke, usually involving several different brain regions. The history is usually punctuated by discrete episodes of abrupt neurological impairment. Mr E gives no such history. Other than excessive alcohol use, hypertension is his only known risk factor for stroke. Of special relevance to his symptoms is diffuse microangiopathy affecting small penetrating cerebral arteries and arterioles. Ischemic changes of this type are reflected in multiple, often confluent areas of abnormal high-signal intensity on T2-weighted or fluid-attenuated inversion recovery MRI (termed leukoaraiosis) involving the periventricular white matter and corona radiata.⁴¹ In such a form of vascular dementia, sometimes called Binswanger disease, cognitive impairment can be insidious in onset and slowly progressive. Motor signs and gait disturbance often accompany the dementia.⁴²

The relationship between alcohol intake and cerebrovascular disease is complex. As with coronary artery disease, mild to moderate alcohol intake appears to reduce the risk of ischemic stroke compared with abstention, whereas heavy intake increases the risk. The relationship thus produces a “J-shaped curve.”¹² A similar J-shaped relationship exists for alcohol consumption and cerebral white matter abnormalities on MRI.⁴³

Dementia can be a feature of Parkinson disease, usually occurring late in the course. With diffuse Lewy body disease, cognitive impairment, often with psychiatric features including hallucinations, can precede symptoms of parkinsonism.⁴⁴ Whether Parkinson disease and diffuse Lewy body disease represent separate entities or a continuum is controversial, but 10% to 30% of older patients with dementia show cortical Lewy bodies at autopsy.^{44 - 45} Mr E's cognitive impairment does not include features that might be expected with diffuse Lewy body disease, and his examination, including his wide-based gait, does not suggest parkinsonism.

Medication effects should be considered in an elderly patient with subtle cognitive symptoms. Mr E takes verapamil and ranitidine, each of which can have cognitive adverse effects,^{46 - 47} and hydrochlorothiazide, which can produce mental symptoms indirectly as a result of hyponatremia or hypercalcemia. None of his medications would be a likely explanation for his memory difficulty, but a temporal correlation between his symptoms and institution of a particular medication or dosage change should be sought.

Treatable dementias in older individuals include a number of metabolic/systemic and primary CNS disorders (Box 3). Relevant to Mr E is cobalamin deficiency, which occurs in approximately 10% of the US population older than 70 years⁴⁸ and can cause neurological symptoms in the absence of anemia, including cognitive change and demyelinating polyneuropathy.⁴⁹ However, for it to do so without evidence of myelopathy would be unusual; moreover, his neurological symptoms did not improve with vitamin B₁₂ treatment.⁵⁰

Metabolic/systemic disorders

Chronic subdural hematoma

Normal pressure hydrocephalus

Chronic central nervous system infection

Brain tumor

Nonconvulsive seizures

Depression

The combination of unsteady gait and altered mentation could suggest normal-pressure hydrocephalus. In that disorder, however, gait disturbance is usually early and prominent, and cognitive impairment more often consists of mental slowing (abulia) than subtle memory impairment.⁵¹

Pseudodementia of depression, at one time considered an artifact of decreased processing speed, appears to be a true cognitive impairment. Nondemented elderly persons who are depressed have reduced hippocampal volumes, and in animal models of depression, hippocampal neurons that normally turn over cease to do so.^{52 - 54} Heavy alcohol use and depression are commonly associated, although alcoholism is more often the cause than the result of the depression.⁵⁵ Mr E does not describe himself as depressed, but he did increase his alcohol intake after retiring, when he was “down in the dumps,” and at times he has been isolated from his family. History-taking in such patients should include screening for depression, including questions about feeling down, insomnia, lack of energy, and poor appetite.⁵⁶

Although pathologically similar and caused by thiamine deficiency, Wernicke and Korsakoff syndromes are clinically distinct. Evolving over days or weeks, Wernicke syndrome consists of the triad of altered mentation, abnormal eye movements, and ataxic gait. Mental abnormalities include lethargy, inattentiveness, abulia, and impaired memory, progressing without treatment to coma. Restricted eye movements progress to complete ophthalmoplegia. Truncal ataxia may prevent standing or walking. Untreated, the condition is fatal, and pathologically there are histologically distinct lesions in the medial thalamus and hypothalamus, the midbrain periaqueductal gray matter, and the pons. Following thiamine and multivitamin replacement, symptoms begin to improve within hours or days, but more than half of treated patients are left with Korsakoff syndrome. In this phase of the disorder, alertness, attentiveness, behavior, and working memory (retention >30 seconds) are relatively preserved, but there is a lasting and often severe amnesia, both anterograde, with inability to retain new information, and retrograde, with lost recall for events months or years old. Insight is often lacking; confabulation may or may not be present. Psychological testing reveals cognitive impairment not explained by pure memory loss, but amnesia predominates.^{57 - 58}

The prevalence of Wernicke-Korsakoff syndrome is uncertain, as Wernicke syndrome is often unrecognized clinically, especially when abnormal eye movements do not accompany altered mentation. In an Australian study, of 131 cases diagnosed at autopsy to have Wernicke encephalopathy, only 26 (20%) had been diagnosed clinically.⁵⁹ In a survey of 22 neuropathologists from Australia, Austria, Belgium, Czechoslovakia, France, Germany, Norway, the United Kingdom, and the United States, the prevalence of Wernicke-Korsakoff syndrome at autopsy ranged from 0% to 2.8%.⁶⁰ Interestingly, Wernicke-Korsakoff prevalence did not correlate with the average amount of alcohol consumed in each country. France, with an average annual per capita alcohol consumption of 12.8 L, had a prevalence of 0.4% to 1.3%. Australia, with average consumption of 8.4 L, had a prevalence of 2.1% to 2.8%.

How often Korsakoff syndrome arises without preceding Wernicke syndrome and how often the mental abnormality is clinically restricted to memory are difficult questions to answer in light of the recognition that ethyl alcohol is directly neurotoxic to both the central and the peripheral nervous systems and that this neurotoxicity can cause cognitive change progressing to dementia in the absence of nutritional deficiency, hepatic encephalopathy, cerebral trauma, or other indirect causes. In animal studies using pair-fed controls, chronic ethanol administration results in neuronal loss in several brain areas, including the hippocampus, the basal forebrain, the cerebral cortex, and the cerebellum.^{61 - 64} A binge pattern of drinking is especially likely to produce brain damage.⁶⁵ The greater damage conferred by repeated bouts of ingestion and withdrawal, compared with continuous drinking, suggests a possible pathophysiological mechanism. Acutely, ethyl alcohol inhibits glutamate excitatory neurotransmission, with rebound-increased glutamatergic transmission during withdrawal.⁶⁶ Repeated bouts of withdrawal might then expose neurons to excitotoxicity and oxidative stress.⁶⁷

In humans, heavy drinking is associated with enlarged ventricles, widened sulci, and abnormal cerebral white matter signals on magnetic resonance imaging.^{68 - 69} Pathologically, neuronal loss occurs in the frontal cerebral cortex, thalamus, hypothalamus, cerebellum, and, less consistently, the hippocampus.⁷⁰ Frontal lobe damage correlates with nonamnestic cognitive abnormalities, such as difficulty planning, organizing, problem solving, and abstracting, as well as perseverative responding and disinhibition.⁷¹

Remarkably, sustained abstinence in such patients can reverse ventricular and sulcal enlargement.^{72 - 73} A study using magnetic resonance spectroscopy showed that morphological improvement correlated with increasing levels of choline (a marker for myelin) in white matter and of N-acetyl-aspartate (a marker for neurons) in frontal cortex. Increased levels of N-acetyl-aspartate were associated with improved cognition.⁷⁴ No one has claimed, however, that cognition returns to normal.

Alzheimer neuropathology and ischemic stroke often coexist in the same patient, and it may then be uncertain to what degree each contributes to cognitive impairment. The task becomes even more difficult when one adds to the equation the adverse effects of heavy ethanol use on cognition and cerebrovascular disease.

If alcohol is a neurotoxin, it is important to define a safe dose threshold. A number of epidemiological studies not only failed to identify a positive relationship between moderate alcohol intake (usually up to 1 or 2 drinks daily) and risk of dementia, but they also found such amounts to be protective.^{75 - 83} Some studies found a special protective effect from wine.^{79 - 80 ,82} Heavier drinking increased the risk of dementia. Alcohol's effects on cognition thus resemble its effects on ischemic vascular disease, producing a J-shaped curve, with low to moderate doses reducing risk and heavy doses increasing risk. Cerebrovascular effects, however, do not fully explain the protective effects of alcohol on cognition. Antioxidant properties of alcoholic beverages have been proposed as a possible mechanism.⁸⁴

Possibly relevant to the apparent benefit of mild to moderate alcohol intake on cognition and ischemic stroke is a recent review challenging alleged protective effects of mild to moderate alcohol on coronary artery disease and all-cause mortality. Surveying 54 epidemiological studies that reported such benefit, the reviewers found that many had compared persons with mild to moderate alcohol intake with a control group consisting of both never drinkers and former drinkers. The authors speculated that former drinkers probably included individuals who stopped drinking because of ill health, and, indeed, when studies using such a comparison group were removed from the analysis, a beneficial effect of mild to moderate alcohol intake was no longer evident.⁸⁵ It remains to be seen whether reports of alcohol's effects on cognition and cerebrovascular disease will withstand similar methodological analysis.

With this information in hand, what are the management options for Mr E? A first step would be to document the degree and nature of his cognitive impairment. Quiz Ref IDThe Mini-Mental State Examination (MMSE) is a 30-point test of cognitive function that assesses orientation, attention, working and recent memory, language, calculation, and copying.⁸⁶ It is insensitive, however, in patients with MCI or early Alzheimer disease in whom subjective concerns about impaired memory may not be reflected in tests of orientation or word recall, and it is not designed to detect depression, executive dysfunction, or behavioral disturbance. In individuals with higher education, score cutoffs at what is considered abnormal are customarily adjusted upward.⁸⁷ The MMSE has acceptable sensitivity and specificity in screening for dementia, but more formal and extensive neuropsychological testing might be necessary to document subtle cognitive dysfunction, especially when impaired memory is not a prominent feature.⁸⁸

The next step would be to exclude other causes of cognitive dysfunction that might be treatable (or even reversible), a task usually accomplished with brain imaging and a few laboratory tests (Box 2). In a review of 9 clinical series comprising 800 individuals aged 65 years or older with dementia, 82 had “treatable” dementias (neurosyphilis, fungal infection, brain tumor, alcohol use, subdural hematoma, hydrocephalus, epilepsy) and 53 had “reversible” dementias (drug toxicity, “metabolic,” hepatic disease, hyponatremia, high or low calcium, vitamin B₁₂ deficiency, thyroid disease, hypoglycemia).⁸⁹ In another series of 200 individuals with dementia, 9.5% had “dementia due to drugs,” 4% “alcohol-related dementia,” 3% hypothyroidism, 1% hyperparathyroidism, 1% hyponatremia, and 0.5% hypoglycemia.⁹⁰

Mr E's eventual working diagnosis will probably be MCI or early Alzheimer disease, an unrelated polyneuropathy, and uncertainty as to the role of alcohol. Given the possibility that alcohol is aggravating symptoms—and it is not unreasonable to speculate that in individuals with MCI or polyneuropathy, a little bit of alcohol might go a longer way; ie, have more harmful effects—abstinence from alcohol should be advised. Mr E may need to participate in a program such as Alcoholics Anonymous or be treated with 1 of the 3 drugs approved by the US Food and Drug Administration for treatment of chronic alcoholism (disulfiram, naltrexone, or acamprosate). If drinking is likely to continue, thiamine and multivitamin supplementation would be appropriate.

As for treating MCI, a double-blind, placebo-controlled study showed that the cholinesterase inhibitor donepezil reduced the likelihood of progression to Alzheimer disease during the first 12 months of the study. The benefit, however, appears to be symptomatic rather than a slowing of disease progression; after 3 years, patients receiving either donepezil or placebo were equally likely to develop Alzheimer dementia. Interestingly, for those carrying the apolipoprotein E ε4 allele, the benefit was evident throughout the 3-year follow-up.⁹¹ Quiz Ref IDCarefully designed studies have also shown that cognitive leisure activities (reading, writing, crossword puzzles, board or card games, group discussions, or playing music, but not physical activities) reduce the risk of developing MCI or dementia in elderly persons.^{92 - 95} (In 1 of these studies, watching television was associated with an increased risk of cognitive impairment.^{96 - 97} ) There is also evidence that specific cognitive training sessions can reduce the risk of cognitive decline. Despite his subjective concerns about memory impairment, Mr E appears to have more than enough mental capacity to engage in activities of this sort.

QUESTION: Why did you exclude alcoholic cerebellar degeneration, or at least put that lower on the list? He was complaining of difficulty with his hands; he couldn't use his hands any more. Eugene O’Neill, the American playwright, was one of Raymond Adams' first 10 cases of alcoholic cerebellar degeneration described from Boston City Hospital. O’Neill complained he could no longer hold a pen, he couldn't write, he couldn't complete his plays, and this was very distressing to him. But you didn't consider this high on the list for this patient?

DR BRUST: I didn't exclude alcoholic cerebellar degeneration; I simply believe the major contributor to Mr E's gait difficulty is his polyneuropathy. The polyneuropathy might be affecting proprioception in his feet, resulting in so-called sensory ataxia, or his unsteadiness might be largely due to weakness. As for his hand weakness, superimposed ulnar palsy was documented by electromyogram and is evident on examination, with clawing of the fourth and fifth digits.

Incidentally, an autopsy report several years ago by Price and Richardson⁹⁸ revealed that Eugene O’Neill probably had a familial or sporadic nonalcohol/nonnutritional cerebellar degenerative disease. During the last 25 years of his life O’Neill did not drink alcohol excessively, and during his last 8 years he was abstinent, yet cerebellar ataxia affecting gait, limbs, and articulation progressed relentlessly from around age 50 years until his death at age 65 years. His arms and hands were affected so severely that he could no longer write. Alcoholic cerebellar degeneration tends to cause truncal ataxia while sparing the limbs, especially the arms; such patients are unlikely to have upper limb tremor (except as a withdrawal phenomenon), dysmetria, or dysdiadochokinesis.⁹⁹

Alcoholic cerebellar degeneration can occur without other signs of Wernicke-Korsakoff syndrome,⁹⁹ and I cannot exclude it as possibly contributing to Mr E's gait difficulty (which until recently was mild enough to allow him to play tennis). However, I don't think we need either alcoholic cerebellar degeneration or, for that matter, alcoholic polyneuropathy to account for his symptoms.

QUESTION: It's very easy for us to say that he should abstain from alcohol, but it's very hard to effect that as a primary caregiver. If we're to believe his report, he had decades of little alcohol intake, followed by what's been a relatively short, perhaps 10-year interval of heavy alcohol intake. Is there a safe level of alcohol for older patients? Does he need to completely abstain or, rather, to lower his intake below some level that is too high?

DR BRUST: This gets into the thorny definition of what is alcoholism. It's a huge spectrum, ranging from people who are physically dependent and get the shakes and delirium tremens when they stop to those who don't drink every day, but when they do drink get into bad trouble. I think, though, that for people who are drinking every day (and he is), it's very risky to try to be a normal social drinker.

There is a phenomenon called sensitization, currently a hot topic in substance abuse research and in animals, and most clearly seen with psychostimulant drugs. Depending on dosage scheduling, a point is reached at which, after a period of abstinence, exposure to the drug triggers an excessive response (“reverse tolerance”) and craving.¹⁰⁰ Such a phenomenon might explain the alcoholic who, abstinent for weeks or months, takes 1 drink and then cannot stop. Alcoholics Anonymous has always recognized this danger and is adamant that problem drinkers should not drink at all. One drink, or even being in an environment associated with drinking, can trigger sensitization and craving—similar to Pavlov's dogs salivating at the sound of the bell.¹⁰¹ So I recommend abstinence, even though it is seldom achieved. Treating alcoholism is one of the toughest jobs we have.

QUESTION: Is there any safe level of alcohol for the aging brain?

DR BRUST: In an aging brain that is having cognitive difficulties? I don't know. It is possible that low doses of alcohol are protective in the normal brain and may even stave off MCI or Alzheimer disease. But if you already have damage, does a little bit of alcohol go the wrong way? I don't know any evidence one way or the other on that.

QUESTION: We tell people with cardiovascular disease that if they quit smoking, they can derive some benefit in as early as 2 years or so, returning their risk to that of a nonsmoker, with respect to an acute cardiovascular event. Can we tell this patient that he might have similar benefit or improvement in his cognitive impairment, were he to quit drinking?

DR BRUST: If his cognitive impairment is entirely attributable to alcohol, he might get better. The study I cited showing improved cognition associated with rising levels of brain N-acetyl-aspartate is very promising.⁷⁴ If, as is probably more likely, he has mild cognitive impairment independent of alcohol (or aggravated by alcohol), he may get a little better, but he has an 80% chance of his MCI progressing to dementia over the next several years.

QUESTION: In terms of patients who have had reversal of MCI, are those patients who have performed cognitive or physical exercises, or is there some other mechanism that's operative?

DR BRUST: I don't know. Studies comparing those who progressed to dementia with those who did not gave group numbers but did not describe individual behavioral patterns. The evidence that cognitive leisure benefits cognition in older people makes it plausible that those who improved kept themselves occupied doing crossword puzzles or playing the clarinet.

QUESTION: I’m a little leery of his story of not drinking a lot when he was an executive in large companies. My experience over the years suggests that when the “2-martini lunch” people get older, they often don't think as crisply as I would expect a person with a high level of executive function to think. Has that been studied over time?

DR BRUST: Higher educational level reduces the risk of developing Alzheimer dementia.¹⁰² Moreover, a study of elderly nuns assessed autobiographic essays they had written more than 50 years earlier for “linguistic ability” based on “idea density and grammatical complexity.” Those who scored lowest on linguistic ability in their 20s were most likely to have low cognitive test scores late in life and, among those who died, to have neuropathologically confirmed Alzheimer disease.¹⁰³ Were plaques already setting in by age 20 years, or did the elderly nondemented nuns have more cognitive reserve when their plaques came along? I don't know of any study that singled out high-powered executives.