Alcohol functions as a central nervous system (CNS) depressant. Some of the early effects of alcohol, such as garrulousness, aggressiveness, excessive activity, and increased electrical excitability of the cerebral cortex—all of them suggestive of cerebral stimulation—are thought to be caused by the inhibition of certain subcortical structures (possibly the high brainstem reticular formation) that ordinarily modulate cerebrocortical activity. Similarly, the initial hyperactivity of tendon reflexes may represent a transitory escape of spinal motor neurons from higher inhibitory centers. With increasing amounts of alcohol, however, the depressant action involves the cortical as well as other brainstem and spinal neurons. All manner of motor functions—whether the simple maintenance of a standing posture, the control of speech and eye movements, or highly organized and complex motor skills—are adversely affected by alcohol. The movements involved in these acts are not only slower than normal but also more inaccurate and random in character and therefore less well adapted to the accomplishment of specific ends.
Alcohol also impairs the efficiency of mental function by interfering with the speed of perception and the ability to persist in mental processing. The learning process is slowed and rendered less effective. Facility in forming associations, whether of words or of figures, and the ability to focus, sustain attention, and concentrate are reduced. Finally, alcohol impairs the faculties of judgment and discrimination and, all in all, the ability to think and reason clearly.
A number of neurologic disorders are characteri-stically associated with alcoholism. The factor common to all of them, of course, is the abuse of alcohol, but the mechanism by which alcohol produces its effects varies widely from one group of disorders to another and in many cases, the essential problem is one of nutritional deficiency as discussed in the preceding chapter. The classification that follows is based for the most part on known mechanisms.
Alcohol Intoxication and Related Disorders
The usual manifestations of alcohol intoxication are so commonplace that they require little elaboration. They consist of varying degrees of exhilaration and excitement, loss of restraint, irregularity of behavior, loquacity and slurred speech, incoordination of movement and gait, irritability, drowsiness, and, in advanced cases, sleepiness, stupor, and coma. There are several complicated types of alcohol intoxication, which are considered below.
As has been indicated, the symptoms of alcoholic intoxication are the result of the depressant action of alcohol on cerebral and spinal neurons. In this respect alcohol acts on nerve cells in a manner akin to the general anesthetics and can cause coma. Unlike the anesthetics, however, the margin between the dose of alcohol that produces surgical anesthesia and that which dangerously depresses respiration is a narrow one, a fact that adds an element of urgency to the diagnosis and treatment of alcoholic coma. One must also be alert to the possibility that other sedative-hypnotic drugs may have potentiated the depressant effects of alcohol. Another treacherous situation is that of traumatic brain injury that is complicated by intoxication, a circumstance that is prone to misinterpretation because of uncertainty as to the main cause of stupor or coma.
Despite what has been said earlier, on rare occasions, alcohol has an exclusively excitatory rather than a sedative effect. This reaction has been referred to in the past as pathologic, or complicated, intoxication and as acute alcoholic paranoid state. Because all forms of intoxication are pathologic, atypical intoxication or idiosyncratic alcohol intoxication are more appropriate designations. Nevertheless, the term pathologic intoxication is still widely used. The boundaries of this syndrome have never been clearly drawn. In the past, variant forms of delirium tremens and epileptic phenomena, as well as psychopathic and criminal behavior, were indiscriminately included. Now the term is generally used to designate an outburst of blind fury with assaultive and destructive behavior. Often the patient is subdued only with difficulty. The attack terminates with deep sleep, which occurs spontaneously or in response to parenteral sedation; on awakening, the patient has no memory of the episode. Lesser degrees are also known wherein the patient, after several drinks, repeatedly commits gross social indiscretions. Allegedly this reaction may follow the ingestion of a small amount of alcohol, but in some of the patients we have observed the amount has often been substantial. Unlike the usual forms of alcohol intoxication and withdrawal, the atypical form has not been produced in experimental subjects, and the diagnosis depends upon the aforementioned arbitrary criteria.
Pathologic intoxication has been ascribed to many factors, but there are no meaningful data to support any of them. However, an analogy may be drawn between pathologic intoxication and the paradoxical reaction that occasionally follows the administration of barbiturates or other sedative drugs. The few patients we have seen, mostly young men of college age or slightly older, have been docile and seemingly well adjusted when not drinking. Usually, they have avoided alcohol after a first episode of this sort, but there have been exceptions.
The main disorders to be distinguished from pathologic intoxication are temporal lobe seizures that occasionally take the form of outbursts of rage and violence and the explosive episodes that characterize the behavior of certain sociopaths. The diagnosis in these cases may be difficult and depends on eliciting the other manifestations of temporal lobe epilepsy or sociopathy. Pathologic intoxication may require the use of restraints and the parenteral administration of diazepam (5 to 10 mg) or haloperidol (2 to 5 mg), repeated once after 30 to 40 min if necessary.
In the language of the alcoholic, the term blackout refers to an interval of time, during a period of severe intoxication, for which the patient later has no memory—even though the state of consciousness, as observed by others, was not grossly altered during that interval. However, a systematic assessment of mental function during the amnesic period has usually not been made. A few observations indicate that it is short-term (retentive) memory rather than immediate or long-term memory that is impaired; this feature and the subsequent amnesia for the episode are vaguely reminiscent of the disorder known as transient global amnesia (see Chap. 21) but without the incessant repetitive questioning and competence in nonmemory mental activities that characterize the latter.
Blackouts may occur at any time in the course of alcoholism, even during the first drinking experience, and they certainly have happened in persons who never became alcoholic. The salient facts are that there is a degree of intoxication that interferes with the registration of events and the formation of memories during the period of intoxication and that the amount of alcohol consumed in moderate social drinking will rarely produce this effect.
Treatment of Severe Alcohol Intoxication
Coma caused by alcohol intoxication represents a medical emergency. The main objective of treatment is to prevent respiratory depression and its complications as described in Chap. 17. One would like to lower the blood alcohol level as rapidly as possible. The previously favored administration of fructose or of insulin and glucose for this purpose is now known to be of little value. Analeptic drugs such as amphetamine and various mixtures of caffeine and picrotoxin are antagonistic to alcohol only insofar as they are overall nervous system excitants but they do not hasten the oxidation of alcohol and are not clinically useful. The use of hemodialysis should be considered in comatose patients with extremely high blood alcohol concentrations (>500 mg/dL), particularly if accompanied by acidosis, and in those who have concurrently ingested methanol or ethylene glycol or some other dialyzable drug.
Methyl, Amyl, and Isopropyl Alcohols and Ethylene Glycol
Poisoning with alcohols other than ethyl alcohol is a rare but catastrophic occurrence. Amyl alcohol (fusel oil) and isopropyl alcohol are used as industrial solvents and in the manufacture of varnishes, lacquers, and pharmaceuticals; in addition, isopropyl alcohol is readily available as a rubbing alcohol. Intoxication may follow the ingestion of these alcohols or inhalation of their vapors. The effects of both are much like those of ethyl alcohol, but much more toxic. They also have in common the generation of acidosis, usually with an anion gap and if a sample of serums obtained soon after the ingestion, an osmolar gap that represents the molecules of the circulating alcohol.
Methyl alcohol (methanol, wood alcohol) is a component of antifreeze and many combustibles and is used in the manufacture of formaldehyde, as an industrial solvent, and as an adulterant of alcoholic beverages, the latter being the most common source of methyl alcohol intoxication. The oxidation of methyl alcohol to formaldehyde and formic acid proceeds relatively slowly; thus, signs of intoxication do not appear for several hours or may be delayed for a day or longer. Many of the toxic effects are like those of ethyl alcohol, but in addition severe methyl alcohol poisoning may produce serious degrees of acidosis (with an anion gap). The characteristic features of this intoxication, however, are damage to retinal ganglion cells—giving rise to scotomata and varying degrees of blindness, dilated unreactive pupils, and retinal edema—and bilateral degeneration of the putamens, readily visible on CT scans. Survivors may be left blind or, less often, with putamenal necrosis and dystonia or Parkinson disease (McLean et al). The most important aspect of treatment is the intravenous administration of large amounts of sodium bicarbonate to reverse acidosis. Hemodialysis and 4-methylpyrazole (see later) may be useful adjuncts because of the slow rate of oxidation of methanol.
Ethylene glycol, an aliphatic alcohol, is a commonly used industrial solvent and the major constituent of antifreeze. In the latter form, it is sometimes consumed by skid-row alcoholics (5,000 cases of poisoning annually in the United States) and in suicide attempts with disastrous results. At first the patient merely appears drunk, but after a period of 4 to 12 h, hyperventilation and severe metabolic acidosis develop, followed by confusion, convulsions, coma, and renal failure and death in rapid succession. Cerebrospinal fluid lymphocytosis is a common but not invariable feature. The metabolic acidosis is a result of the conversion of ethylene glycol by alcohol dehydrogenase into glycolic acid, thus producing an anion gap that reflects the presence of this additional substance in the blood. (The anion gap has been defined in different ways, but the most convenient is the difference between the positive ion Na+ and the sum of negative ions, Cl– plus HCO3– [venous CO2 is used for the latter]; a value greater than 12 is considered a gap.) The cause of the renal toxicity is less clear—probably it is a result of the formation of oxalate from glycolate and the deposition of oxalate crystals in renal tubules. (One of our recent patients had hippurate crystals in the urine, a finding that is more characteristic of toluene ingestions.) These crystals appear in the urine and sometimes in the cerebrospinal fluid (CSF) and aid in diagnosis.
Treatment of Nonethanol Alcohol Intoxication
The treatment of ethylene glycol poisoning has, until relatively recently, consisted of hemodialysis and the intravenous infusion of sodium bicarbonate and ethanol, the latter serving as a competitive substrate for alcohol dehydrogenase. However, the use of ethanol in this regimen is problematic. Baud and colleagues, and more recently Brent and colleagues and Jacobsen, have advocated the use of intravenous 4-methylpyrazole (fomepizole), which is a far more effective inhibitor of alcohol dehydrogenase than is alcohol. They recommend this form of treatment for methanol poisoning as well. Information from the American Academy of Toxicology is cited in a review of the use of fomepizole by Brent, which is recommended to the interested reader. Generally, for either methanol of ethylene glycol, a plasma level of the alcohol above 20 mg/dL, or above 10 mg/dL when combined with an osmolal gap over 10 is considered appropriate to institute the drug. In the case of ethylene glycol, oxaluria and acidosis are additional factors that may precipitate treatment. Dialysis remains an essential therapy if cerebral and renal damage is not too advanced.
Some patients who recover from the acute renal and metabolic effects are left with multiple cranial nerve defects, particularly of the seventh and eighth nerves. The latter abnormalities develop 6 to 18 days after the ingestion of ethylene glycol and have been attributed to the deposition of oxalate crystals along the subarachnoid portions of the affected nerves (Spillane et al).
The Alcohol Abstinence, or Withdrawal, Syndrome
This is the well-known symptom complex of tremulousness, hallucinations, seizures, confusion, and psychomotor and autonomic overactivity. Although a sustained period of chronic inebriation is the most obvious factor in the causation of these symptoms, they become manifest only after a period of relative or absolute abstinence from alcohol—hence the designation abstinence, or withdrawal, syndrome. Figure 42-1 illustrates this concept. Each of the major manifestations of the withdrawal syndrome may occur in more or less pure form and are so described below, but usually they occur in combination. Major withdrawal symptoms are observed mainly in the binge, or periodic, drinker, although the steady drinker is not immune if for some reason he stops drinking, such as during a hospital admission for surgery or a medical illness. The full syndrome, depicted further on, is called delirium tremens.
Relation of acute neurologic disturbances to cessation of drinking. The shaded drinking period is greatly foreshortened and not intended to be quantitative. The periodic notching in the baseline represents the tremulousness, nausea, and so on that occur following a night's sleep. The time relations of the various groups of symptoms to withdrawal are explained in the text. (Adapted from Victor M, Adams RD: The effect of alcohol on the nervous system. Res Publ Assoc Res Nerv Ment Dis 32:526, 1953, by permission.)
The most common single manifestation of the abstinence syndrome is tremulousness, often referred to as "the shakes" or "the jitters," combined with general irritability and gastrointestinal symptoms, particularly nausea and vomiting. These symptoms first appear after several days of drinking, usually in the morning after a night's abstinence. The patient "quiets his nerves" with a few drinks and is then able to drink for the rest of the day without undue distress. The symptoms return on successive mornings with increasing severity. The symptoms then become augmented, reaching their peak intensity 24 to 36 h after the complete cessation of drinking. Generalized tremor is the most obvious feature. It is of fast frequency (6 to 8 Hz), slightly irregular, and variable in severity, tending to diminish when the patient is in quiet surroundings and to increase with motor activity or emotional stress. The tremor may be so violent that the patient cannot stand without help, speak clearly, or eat without assistance. Sometimes there is little objective evidence of tremor, and the patient complains only of being "shaky inside."
Within a few days, flushed facies, anorexia, tachycardia, and tremor characteristic of the mild withdrawal syndrome subside to a large extent, but overalertness, tendency to startle easily, and jerkiness of movement may persist for a week or longer. Feelings of uneasiness may not leave the patient completely for 10 to 14 days. According to Porjesz and Begleiter, certain electrophysiologic abnormalities (diminished amplitudes of sensory evoked potentials and prolonged latencies and conduction velocities of auditory brainstem potentials) remain altered long after the clinical abnormalities have subsided.
Symptoms of disordered perception occur in about one-quarter of withdrawing hospitalized tremulous patients. The patient may complain of "bad dreams"—nightmarish episodes associated with disturbed sleep—which he finds difficult to separate from real experience. Sounds and shadows may be misinterpreted, or familiar objects may be distorted and assume unreal forms (illusions). There may also be more overt hallucinations, which are purely visual in type, mixed visual and auditory, tactile, or olfactory, in this order of frequency. There is little evidence to support the popular belief that certain visual hallucinations (bugs, pink elephants) are specific to alcoholism. Actually, the hallucinations comprise the full range of visual experience. They are more often animate than inanimate; persons or animals may appear singly or in panoramas, shrunken or enlarged, natural and pleasant, or distorted, hideous, and frightening. The hallucinosis may be an isolated phenomenon lasting for a few hours, and it may later be attended by other withdrawal signs.
Acute and Chronic Auditory Hallucinosis
A special type of alcoholic psychosis consisting of a more or less pure auditory hallucinosis has been recognized for many years. Kraepelin referred to it as the "hallucinatory insanity of drunkards," or "alcoholic mania." A report of 75 such cases was made by Victor and Hope. The central feature of the illness, in the beginning, is the occurrence of auditory hallucinations despite an otherwise clear sensorium during the withdrawal period; i.e., the patients are not disoriented or obtunded, and they have an intact memory. The hallucinations may take the form of unstructured sounds such as buzzing, ringing, gunshots, or clicking (the elementary hallucinations of Bleuler), or they may have a musical quality, like a low-pitched hum or chant. The most common hallucinations, however, are human voices. When the voices can be identified, they are often attributed to the patient's family, friends, or neighbors—rarely to God, radio, or television. The voices may be addressed directly to the patient, but more frequently, they discuss him in the third person. In the majority of cases, the voices are maligning, reproachful, or threatening in nature and are disturbing to the patient; a significant proportion, however, are not unpleasant and leave the patient undisturbed. To the patient, the voices are clearly audible and intensely real, and they tend to be exteriorized; i.e., they come from behind a radiator or door, from the corridor, or through a wall, window, or floor. Another feature of auditory hallucinosis is that the patient's response is more or less understandable in light of the hallucinatory content. The patient may call on the police for protection or erect a barricade against invaders; he may even attempt suicide to avoid what the voices threaten. The hallucinations are most prominent during the night, and their duration varies greatly: they may be momentary, or they may recur intermittently for days on end and, in exceptional instances, for weeks or months.
While hallucinating, most patients have no appreciation of the unreality of their hallucinations. With improvement, the patient begins to question the inauthenticity and may be reluctant to talk about them and may even question his own sanity. Full recovery is characterized by the realization that the voices were imaginary and by the ability to recall, sometimes with remarkable clarity, some of the abnormal thought content of the psychotic episode.
A unique feature of this alcoholic psychosis is its evolution, in a small proportion of the patients, to a state of chronic auditory hallucinosis. The chronic disorder begins like the acute one, but after a short period, perhaps a week or two, the symptomatology begins to change. The patient becomes quiet and resigned, even though the hallucinations remain threatening and derogatory. Ideas of reference and influence and other poorly systematized paranoid delusions become prominent. At this stage the illness may be mistaken for paranoid schizophrenia and indeed was so identified by Bleuler. There are, however, important differences between the two disorders: the alcoholic illness develops in close relation to a drinking bout and the past history rarely reveals schizoid personality traits. Moreover, alcoholic patients with hallucinosis are not distinguished by a high incidence of schizophrenia within their families (Schuckit and Winokur; Scott), and a large number of such patients, whom our colleagues Victor and Adams evaluated long after their acute attacks, did not show signs of schizophrenia. There is some evidence that repeated attacks of acute auditory hallucinosis render the patient more susceptible to the chronic state.
Withdrawal Seizures ("Rum Fits")
In the setting of alcohol withdrawal either as relative or absolute abstinence following a period of chronic inebriation, convulsive seizures are common. More than 90 percent of withdrawal seizures occur during the 7- to 48-h period following the cessation of drinking, with a peak incidence between 13 and 24 h. During the period of seizure activity, the electroencephalogram (EEG) is usually abnormal, but it reverts to normal in a matter of days, even though the patient may go on to develop delirium tremens. During the period of seizure activity and for days afterwards, the patient is unusually sensitive to stroboscopic stimulation; almost half the patients respond with generalized myoclonus or a convulsive seizure (photoparoxysmal response).
Seizures occurring in the abstinence period have a number of other distinctive features. There may be only a single seizure, but in the majority of cases the seizures occur in bursts of 2 to 6 over a day, occasionally even more; only 2 percent of patients studied by Victor (1968) developed status epilepticus. The seizures are grand mal in type, i.e., generalized, tonic-clonic convulsions with loss of consciousness. Focal seizures should always suggest the presence of a focal brain lesion (most often traumatic) in addition to the effects of alcohol. Twenty-eight percent of Victor's patients with generalized withdrawal seizures went on to develop delirium tremens (the percentage has been less in other series); almost invariably, the seizures preceded the delirium. The postictal confusional state may blend imperceptibly with the onset of the delirium, or the postictal state may have cleared over several hours or even a day or longer before the delirium sets in. Seizures of this type typically occur in a patient whose drinking history has extended over a period of many years and must be distinguished from other forms of seizures that have their onset in adult life. The term rum fits, or whiskey fits—the names sometimes used by alcoholics—is reserved for seizures with the attributes described here. This serves to distinguish them from seizures that occur in the interdrinking period, long after withdrawal has been accomplished.
It is important to note that the common idiopathic or posttraumatic forms of epilepsy are also influenced by alcohol. In these types of epilepsy, a seizure or seizures may be precipitated by only a short period of drinking (e.g., a weekend, or even one evening of heavy social drinking); perhaps unsurprisingly in these circumstances, the seizures occur not when the patient is intoxicated but usually the morning after, in the "sobering-up" period. Except for the transient dysrhythmia in the withdrawal period, the incidence of EEG abnormalities in patients who have had rum fits is no greater than in normal persons, in sharp contrast to the EEGs of nonalcoholic patients with recurrent seizures.
Treatment and Prevention of Withdrawal Seizures
Most patients during withdrawal do not require antiepileptic drugs, as the entire episode of seizure activity—whether a single seizure or a brief flurry of seizures—may have terminated before the patient is brought to medical attention. The parenteral administration of diazepam or sodium phenobarbital early in the withdrawal period does, however, prevent withdrawal fits in patients with a previous history of this disorder, as well as in those who might be expected to develop seizures on withdrawal of alcohol. This approach has been supported by the observations of D'Onofrio and colleagues that intravenous lorazepam (2 mg in 2 mL of normal saline) was highly effective in preventing recurrent seizures after a first seizure in the same withdrawal period. Only 3 of 100 patients so treated had a second seizure within 48 h, compared to 21 of 86 untreated patients. The long-term administration of anticonvulsants is neither necessary nor practical: if such patients remain abstinent, they will be free of seizures; if they resume drinking, they usually abandon their medications. Furthermore, it is not certain that continued administration of anticonvulsants dependably prevents abstinence seizures. The rare instances of status epilepticus should be managed like status of any other type (see Chap. 16). In alcoholics with a history of idiopathic or posttraumatic epilepsy, the goal of treatment should be abstinence from alcohol, because of the tendency of even short periods of drinking to precipitate seizures. Such patients need to be maintained on anticonvulsant drugs.
Delirium Tremens and Related Disorders
This is the most dramatic and grave of all the acute alcoholic illnesses. It is characterized by profound confusion, delusions, vivid hallucinations, tremor, agitation, and sleeplessness, as well as by the signs of increased autonomic nervous system activity—i.e., dilated pupils, fever, tachycardia, and profuse perspiration. The clinical features of delirium are presented in detail in Chap. 20 as they relate to delirium tremens (DTs) and to other illnesses that simulate it.
Delirium tremens develops in one of several settings. The patient, an excessive and steady drinker for many years, may have been admitted to the hospital for an unrelated illness, accident, or operation and, after 2 to 4 days, occasionally even later, becomes delirious. Or, following a prolonged drinking binge, the patient may have experienced several days of tremulousness and hallucinosis or one or more seizures and may even be recovering from these symptoms when delirium tremens develops, rather abruptly as a rule.
As to the frequency of delirium tremens, Foy and Kay reported an incidence of 0.65 percent of all patients admitted for other reasons to a large general hospital. Among 200 consecutive alcoholics admitted to a city hospital, Ferguson and colleagues reported that 24 percent developed delirium tremens; of these, 8 percent died—figures that are considerably higher than those recorded in our hospitals (see below). Of course, the reported incidence of delirium tremens will vary greatly, depending on the population served by a particular hospital.
In the majority of cases, delirium tremens is benign and short-lived, ending as abruptly as it begins. Consumed by relentless activity and wakefulness for several days, the patient falls into a deep sleep and then awakens lucid, quiet, and exhausted, with virtually no memory of the events of the delirious period. Less commonly, the delirious state subsides gradually with intermittent episodes of recurrence. In either event, when delirium tremens occurs as a single episode, the duration is 72 h or less in more than 80 percent of cases. Less frequently still, there may be one or more relapses, several episodes of delirium of varying severity being separated by intervals of relative lucidity—the entire process lasting for several days or occasionally for as long as 4 to 5 weeks.
In the past, approximately 15 percent of cases of delirium tremens ended fatally, but the figure now is closer to 5 percent. In many of the fatal cases there is an associated infectious illness or injury, but in others, no complicating illness is discernible. Many of the patients die in a state of hyperthermia; in some, death comes so suddenly that the nature of the terminal events cannot be determined. Reports of series of cases with a negligible mortality rate in delirium tremens can usually be traced to a failure to distinguish between delirium tremens and the minor forms of the withdrawal syndrome, which are far more common and practically never fatal.
We make note here of our experience with delirium following the withdrawal of barbiturates (Romero et al), which is almost identical to the DTs, including the abrupt cessation of symptoms, as discussed in the section on "Barbiturate Abstinence, or Withdrawal, Syndrome" in Chap. 43.
There are also alcohol withdrawal states, closely related to delirium tremens and about as frequent, in which one facet of the delirium tremens complex assumes prominence, to the virtual exclusion of the other symptoms. The patient may simply exhibit a transient state of quiet confusion, agitation, or peculiar behavior lasting several days or weeks. Or there may be a vivid hallucinatory–delusional state and abnormal behavior consistent with the patient's false beliefs. Unlike typical delirium tremens, the atypical states usually present as a single circumscribed episode without recurrences, are only rarely preceded by seizures, and do not end fatally.
Pathologic examination is singularly unrevealing in patients with delirium tremens. Edema and brain swelling have been absent in the authors' pathologic material except when shock or hypoxia had occurred terminally. There have been no significant light microscopic changes in the brain, which is what one would expect in a disease that is essentially reversible. The EEG findings have been discussed in relation to withdrawal seizures.
Pathogenesis of the Tremulous-Hallucinatory-Delirious Disorders
Prior to 1950, it was the common belief that these symptoms represented the most severe forms of alcohol intoxication—an idea that fails to satisfy the simplest clinical logic. The symptoms of toxicity—consisting of slurred speech, uninhibited behavior, staggering gait, stupor, and coma—are in themselves distinctive and, in a sense, the opposite of the symptom complex of tremor, fits, and delirium. It is evident, from observations in both humans and experimental animals, that the most important and the one obligate factor in the genesis of delirium tremens and related disorders is the withdrawal of alcohol following a period of sustained chronic intoxication. Furthermore, the emergence of withdrawal symptoms depends on a rapid decline in the blood alcohol level from a previously higher level and not necessarily upon the complete disappearance of alcohol from the blood.
The mechanisms by which the withdrawal of alcohol produces symptoms are incompletely understood. In all but the mildest cases, the early phase of alcohol withdrawal is attended by a drop in serum magnesium concentration and a rise in arterial pH—the latter on the basis of respiratory alkalosis (Wolfe and Victor). Possibly the compounded effect of these two factors, both of which are associated with hyperexcitability of the nervous system, is responsible in part for seizures and for other symptoms that characterize the early phase of withdrawal. However, these factors alone are not explanatory. The molecular mechanisms that are thought to be operative in the genesis of alcohol tolerance and withdrawal have been mentioned earlier. The gabaergic system has been most strongly implicated, in part because the receptors for this inhibitory transmitter are downregulated by chronic alcohol use, but the situation is not nearly so simple, insofar as the excitatory glutaminergic system is also inhibited by alcohol.
Rarely, blood glucose is seriously depressed in the alcohol withdrawal states. Ketoacidosis with normal blood glucose is another infrequent finding. Disturbances of electrolytes are of varying frequency and significance. Serum sodium levels are altered infrequently and are more often increased than decreased. The same is true for chloride and phosphate. Serum calcium and potassium are lowered in about one-quarter of patients. Most patients show some degree of hypomagnesemia, low PCO2, and high arterial pH—abnormalities that are probably important in the pathogenesis of withdrawal symptoms (see later). Abnormalities of the CSF occur unpredictably (it is usually normal), as do changes on CT scanning or MRI; they may indicate the presence of some medical or surgical complication. Enlargement of the third and lateral ventricles is a common finding (see later). The MRI is normal unless there is an incipient Wernicke disease, in which case lesions in the periaqueductal region and subthalami may be evident, as described in the previous chapter.
Treatment of Delirium Tremens and Minor Withdrawal Symptoms (see Chap. 20)
The treatment of delirium tremens begins with a search for associated injuries (particularly head injury with cerebral lacerations or subdural hematoma), infections (pneumonia or meningitis), pancreatitis, and liver disease. Because of the frequency and seriousness of these complications, chest films and a CT scan should be obtained in most instances, and lumbar puncture should be performed if there is suspicion of meningitis. In severe forms of delirium tremens, the temperature, pulse, and blood pressure should be measured at frequent intervals in anticipation of peripheral circulatory collapse and hyperthermia, which, added to the effects of injury and infection, are the usual causes of death in this disease. In the case of hypotension, one must act quickly, using intravenous fluids and, if called for, vasopressor drugs. The occurrence of hyperthermia demands the use of a cooling mattress or evaporative cooling in addition to specific treatment for any infection that may be present.
An additional important element in treatment is the correction of fluid and electrolyte imbalance, particularly hypokalemia and severe hypomagnesemia. Severe degrees of agitation and perspiration may require the administration of up to 5 L of fluid daily, of which at least 1,500 to 2,000 mL should be normal saline. The specific electrolytes and the amounts that must be administered are governed by the laboratory values for these electrolytes. If the serum sodium is extremely low, one must be cautious in raising the level lest a central pontine myelinolysis be induced (see Chap. 40). In the rare case of hypoglycemia, the administration of glucose is an urgent matter. Patients who present with severe alcoholic ketoacidosis and normal or only slightly elevated blood glucose concentrations usually recover promptly, without the use of insulin.
It must be emphasized, as it was in Chap. 41, that a special danger attends the use of glucose solutions in alcoholic patients. The administration of intravenous glucose may serve to consume the last available reserves of thiamine and precipitate Wernicke disease. Typically, these patients have subsisted on a diet disproportionately high in carbohydrate (in addition to alcohol, which is metabolized entirely as carbohydrate) and low in thiamine, and their body stores of B vitamins may have been further reduced by gastroenteritis and diarrhea. For this reason it is good practice to add B vitamins, specifically thiamine (which may also be supplemented by intramuscular injection), in all cases requiring parenterally administered glucose—even though the alcoholic disorder under treatment, e.g., delirium tremens, is not primarily caused by vitamin deficiency.
With respect to the use of medications to treat the withdrawal syndromes, it is important to distinguish between mild symptoms, which are essentially benign and responsive to practically any sedative drug, and full-blown delirium tremens. There is no certain way to predict whether a patient with the early signs of withdrawal will progress to delirium tremens. In the latter state, the object of therapy is to blunt the psychomotor and autonomic overactivity, prevent exhaustion, and facilitate the administration of parenteral fluid and nursing care; one should not attempt to suppress agitation "at all costs," as doing so requires an amount of drug that might depress respiration.
A wide variety of drugs are effective in controlling withdrawal symptoms. The more popular ones have been chlordiazepoxide (Librium), diazepam (Valium), and the ancillary medications, clonidine and beta-adrenergic blockers, and a number of both older and newer anticonvulsant drugs such as gabapentin, which may reduce the requirement for sedative drugs. There is little to choose among the primary sedative drugs in respect to their therapeutic efficacy. More importantly, there are few data to indicate that any one of them can prevent hallucinosis or delirium tremens, or shorten the duration, or alter the mortality rate of the latter disorder (Kaim et al). A contemporary summary of the medication management of withdrawal has been given by Kosten and O'Connor. In general, phenothiazine drugs should be avoided because they may reduce the threshold to seizures. Probably, the use of any of the diazepine medications is as effective as a single dose of lorazepam in prophylactically suppressing seizures (see earlier discussion).
If parenteral medication is necessary, we still prefer 10 mg of diazepam or chlordiazepoxide given intravenously and repeated once or twice at 20- to 30-min intervals until the patient is calm but awake; we also favor midazolam in closely controlled circumstances when hyperactivity and hallucinosis are extreme. Beta-adrenergic–blocking agents, such as propranolol, labetalol, and atenolol, are helpful in reducing heart rate, blood pressure, and the tremor to some extent. Lofexidine, an alpha2-agonist that blocks autonomic outflow centrally, and clonidine are similarly effective in reducing the severity of most of the withdrawal symptoms, but they are not recommended as the sole treatments. Corticosteroids have no place in the treatment of the withdrawal syndrome and more potent agents such as propofol are usually not necessary.
Wernicke-Korsakoff Syndrome and Alcoholic–Nutritional Diseases (see Chap. 41)
Alcoholism provides the ideal setting for the development of nutritional diseases of the nervous system. Although only a small proportion of alcoholics develop nutritional diseases, the overall number of these diseases is substantial because of the frequency of alcoholism. The importance of the alcohol-induced deficiency diseases relates to the fact that they are preventable and, if neglected, may lead to permanent disability. These illnesses, particularly the Wernicke-Korsakoff syndrome, are discussed fully in Chap. 41. Contrary to popular opinion with regard to the prevention of Wernicke disease, the content of B vitamins in American beer and other liquors is so low as to have little nutritional value (Davidson).
Disorders of Uncertain Pathogenesis Associated with Alcoholism
Also discussed in Chap. 41 are alcoholic cerebellar degeneration and Marchiafava-Bignami disease. The former is almost certainly of nutritional origin; in the latter a nutritional–metabolic etiology seems likely but has not been established. Central pontine myelinolysis, although frequently observed in alcoholics, is more appropriately considered with the acquired metabolic disorders, usually the too rapid correction of hyponatremia (see Chap. 40). Certain disorders of skeletal and cardiac muscle associated with alcoholism (acute alcoholic myopathy and cardiomyopathy) are described in Chap. 51, with the myopathies caused by drugs and toxins. There remain to be discussed several diverse disorders that have been attributed to alcoholism but whose causal relationship to alcohol abuse, nutritional deficiency, or some other relevant factor is not clear.
Alcoholic Dementia and Cerebral Atrophy
The term alcoholic dementia is used widely and often indiscriminately to designate a presumably distinctive form of dementia that is attributable to the chronic, direct effects of alcohol on the brain. Unfortunately, a syndrome subsumed under the title of alcoholic dementia and its many synonyms that appear in the older literature (alcoholic deteriorated state, chronic alcoholic psychosis, chronic or organic brain syndrome due to alcohol) has never been delineated satisfactorily, either clinically or pathologically. In the Comprehensive Textbook of Psychiatry, it has been defined as "a gradual disintegration of personality structure, with emotional lability, loss of control, and dementia" (Sadock and Sadock). Purported examples of this state show a remarkably diverse group of symptoms, including jealousy and suspiciousness; coarsening of moral fiber and other personality and behavioral disorders; deterioration of work performance, personal care, and living habits; and disorientation, impaired judgment, and defects of intellectual function, particularly of memory.
There have been many attempts to redefine alcoholic dementia. Cutting, as well as Lishman, expressed the view that the term Korsakoff psychosis should be limited to patients with a fairly pure disorder of memory of acute onset and that patients with more global symptoms of intellectual deterioration, of gradual evolution, be considered to have alcoholic dementia. These are rather weak diagnostic criteria. As pointed out in Chap. 41, Korsakoff psychosis may have an insidious onset and gradual progression, and patients with this disorder, in addition to an amnesic defect, characteristically show disturbances of cognitive functions that depend little or not at all on memory. More importantly, in none of the patients designated by these authors as having alcoholic dementia was there a neuropathologic examination, without which the clinical assessment must remain arbitrary and imprecise.
The pathologic changes that purportedly underlie primary alcoholic dementia are even less precisely defined than the clinical syndrome. Courville, whose writings have been quoted most frequently in this respect, described a series of cerebral cortical changes that he attributed to the toxic effects of alcohol. Some of them turn out on close inspection to be quite nonspecific, reflecting nothing more than the effects of aging or the insignificant artifacts of tissue fixation and staining. Harper and Blumbergs, and subsequently Harper and Kril, reported that the mean brain weight is decreased in alcoholics and the pericerebral space is increased in volume—findings that do no more than confirm the brain shrinkage that is demonstrable by CT scans in many alcoholics and is to some extent reversible with sustained abstinence (see later).
The majority of cases that come to autopsy with the label of alcoholic dementia prove simply to have the lesions of the Wernicke-Korsakoff syndrome. Traumatic lesions of varying degrees of severity are commonly added. Other cases show the lesions of Marchiafava-Bignami disease, hepatic encephalopathy, subdural hematomas, or an unrelated communicating hydrocephalus, Alzheimer disease, ischemic necrosis, or some other disease quite unrelated to alcoholism. Practically always in our material, the clinical state can be accounted for by one or a combination of these disease processes and there has been no need to invoke a hypothetical toxic effect of alcohol on the brain. This has also been the experience of Torvik and associates; with a few exceptions, such as coincidental Alzheimer disease, all their cases that had been diagnosed as having alcoholic dementia turned out, on neuropathologic examination, to have the chronic lesions of Wernicke-Korsakoff disease.
In brief, the most serious flaw in the concept of a primary alcoholic dementia is that it lacks a distinctive, well-defined pathology. Until such time as the morphologic basis is established, its status must remain ambiguous. A more detailed discussion of this subject and of so-called alcoholic cerebral atrophy (see later) can be found in the review by Victor (1994).
Alcoholic cerebral atrophy likewise does not constitute a well-defined entity. The concept was the product originally of pneumoencephalographic studies. Relatively young alcoholics, some with and some without symptoms of cerebral disease were often found to have enlarged cerebral ventricles and widened sulci, mainly of the frontal lobes (see, e.g., reports of Brewer and Perrett and of Haug). Similar findings have been reported in chronic alcoholics examined by CT scanning and MRI (see review by Carlen et al). The clinical correlates of these radiologic findings are unclear. Wilkinson demonstrated that in clinically normal alcoholics, the radiologic measures of "brain atrophy" were age related; once the age factor was removed, the CT findings in these subjects did not differ significantly from those in nonalcoholic controls. However, from the studies by Harper and colleagues (1982 and 1985) it may indeed be the case that chronic exposure to alcohol induces cerebral atrophy, but this requires confirmation. The idea of alcoholic atrophy is open to criticism mainly on the grounds that dilated ventricles have in fact been reversible to a considerable extent when abstinence is maintained (Carlen et al; Lishman; Zipursky et al; Schroth et al).
That parental alcoholism may have an adverse effect on the offspring has been a recurrent theme in medical writings. Probably the first allusion to such a relationship was that of Sullivan in 1899, who reported that the mortality among the children of drunken mothers was more than two times greater than that among children of nondrinking women of "similar stock." This increased mortality was attributed by Sullivan and later by Haggard and Jellinek to postnatal influences such as poor nutrition and a chaotic home environment, rather than to the intrauterine effects of alcohol. The idea that maternal alcoholism could damage the fetus was generally rejected and relegated to the category of superstitions about alcoholism or the claims of temperance ideologues.
In the late 1960s, the effects of alcohol abuse on the fetus were rediscovered, so to speak. Lemoine and associates in France, and then Ulleland and Jones and Smith in the United States, described a distinctive pattern of abnormalities in infants born of severely alcoholic mothers. They stated that the affected infants are small in length in comparison to weight, and most of them fall below the third percentile for head circumference. They are distinguished also by the presence of short palpebral fissures (shortened distance between inner and outer canthi) and epicanthal folds; maxillary hypoplasia, micrognathia, indistinct philtrum, and thin upper lip; and longitudinally oriented palmar creases, flexion deformities of the fingers, and a limited range of motion of other joints. Minor anomalies (usually spontaneously closing cardiac septal defects), anomalous external genitalia, and cleft lip and palate are much more frequent than in the general population. All of these features have similarities to the syndrome described in a proportion of infants whose mothers had taken anticonvulsants during pregnancy, the "fetal anticonvulsant syndrome" (see Chap. 16). The newborn infants suck and sleep poorly, and many of them are irritable, restless, hyperactive, and tremulous; these last symptoms resemble those of alcohol withdrawal except that they persist.
The first long-term study of children with what has come to be called fetal alcohol syndrome (FAS) was reported by Jones and coworkers. Among 23 infants born to alcoholic mothers, there was a neonatal mortality of 17 percent; among the infants who survived the neonatal period, almost half failed to achieve normal weight, length, and head circumference or remained backward mentally to a varying degree, even under optimal environmental conditions. Several large groups of severely affected children have now been observed for 20 years or longer (see Streissguth). Distractibility, inattentiveness, hyperactivity, and impairment of fine motor coordination are prominent features in early childhood. Most such children fall into the category of attention-deficit hyperactivity disorder. Slow growth of head circumference is a consistent finding throughout infancy and childhood. The physical stigmata of the syndrome become less distinctive after puberty, but practically all adolescents are left with some degree of mental retardation and behavioral abnormalities.
The pathologic changes that underlie the syndrome have been studied in a small number of cases and no uniform change has emerged. Of some interest are observations such as those of Ikonomidu and coworkers that demonstrate a profound effect of alcohol exposure on the deletion of millions of neurons in the developing rat brain by a mechanism of apoptosis. The main vulnerability occurs during periods of synaptogenesis, which in humans extends from the sixth month of gestation onward.
It is noteworthy that infants born to nonalcoholic mothers who had been subjected to severe dietary deprivation during pregnancy (during World War II) were small and often premature, but these infants did not show the pattern of malformations that characterizes FAS. Alcohol readily crosses the placenta in humans and animals; in the mouse, rat, chick, miniature swine, and beagle dog, alcohol has been shown to have both embryotoxic and teratogenic effects. Thus, the evidence to date favors a toxic effect of alcohol, although a possible toxic effect of acetaldehyde and smoking and a possible contributory role for nutritional deficiency have not been totally excluded.
Unequivocal cases of FAS observed to date have occurred only in infants born to severely alcoholic mothers (many of them with delirium tremens and liver disease) who continued to drink heavily throughout pregnancy. It is important to state that a relationship to lesser degrees of alcohol intake is far less secure. Data derived from the collaborative study sponsored by the National Institutes of Health indicate that about one-third of the offspring of women who are heavy drinkers have FAS. Abel and Sokol have estimated that the worldwide incidence of FAS is 1.9 per 1,000 live births and have pronounced it the leading known cause of mental retardation in the western world. The degree of maternal alcoholism that is necessary to produce the syndrome and the critical stage in gestation during which it occurs are still vague. The various teratogenic effects described earlier are estimated to occur in the embryonic period, i.e., in the first 2 months of fetal life. Other nonteratogenic effects appear to be related to periods during gestation when the fetus is exposed to particularly high alcohol levels.
A comprehensive and not outdated account of alcohol-related birth defects and the controversial issues surrounding this subject is contained in a special issue of Alcohol Health and Research World, published by the National Institutes of Health (Vol. 18, 1994).
Neurologic Complications of Alcoholic Cirrhosis and Portal–Systemic Shunts
This category of alcoholic disease is discussed in Chap. 40, in connection with the acquired metabolic disorders of the nervous system.