Alcoholic and Other Toxic Dementias Eileen M. Joyce

ALCOHOLIC DEMENTIA

The concept of a dementia consequent upon the neurotoxic effects of long-term alcohol abuse developed from clinical observations of a gradual deterioration in personality and intellect in many alcoholics. Subsequent research suggested that this is age related, is milder in degree than other forms of dementia and constitutes 11–24% of all demented patients^1,2. To put this in further perspective, of the estimated 18 million alcoholics in the USA, approximately 40% will have mild cognitive impairment which is potentially reversible with abstinence, but a further 10% will have persisting severe cognitive impairment requiring institutional care, comprising patients mainly with dementia and amnesia^3,4.

Amnesia in alcoholics is known as the Wernicke-Korsakoff syn-drome. This is the result of thiamine malnutrition which causes subcortical periventricular vascular lesions in diencephalic structures, the mamillary body and thalamus, known to be critically involved in new memory formation⁵. Alcoholic dementia was originally thought to be distinct from the Wernicke-Korsakoff syndrome and due to the direct neurotoxic effect of alcohol on the cerebral cortex. Horvath⁶ prospectively examined 100 chronic alcoholics presenting clinically with a dementia syndrome and concluded that alcoholic demen-tia was different from the Wernicke-Korsakoff syndrome and that other non-amnesic organic syndromes exist in alcoholics characteris-tic of frontal, parietal or global cortical damage. Cutting⁷ performed a retrospective case-note analysis of alcoholic patients with cogni-tive impairment and found two forms of clinical presentation. One consisted of a rapidly developing illness in younger patients with preserved intellect, akin to the traditional Wernicke-Korsakoff syn-drome, whereas the other was more characteristic of dementia, being a gradual and global cognitive deterioration in older patients.

Although these studies support the concept of alcoholic dementia as a distinct entity, other evidence suggests a more complex pic-ture with there being several different neuropathological ‘routes’ to the development of brain damage sufficient to cause the dif-fuse cognitive impairment recognized clinically as dementia. Torvik and colleagues⁸ examined the clinical records of patients who at autopsy had diencephalic lesions characteristic of thiamine malnu-trition; that is, Wernicke-type lesions. Seventy five percent of these were considered to be demented in life rather than amnesic and the majority had no additional neuropathological hallmarks of a neurode-generative dementia. They concluded that the diencephalic lesions of thiamine deficiency can result in the clinical picture of both ‘alcoholic dementia’ and the Wernicke-Korsakoff syndrome. Victor and Adams⁹ then pointed out that 10% of their pathologically proven cases of the Wernicke-Korsakoff syndrome developed cognitive abnormali-ties insidiously rather than acutely⁵ and that cognitive impairment other than amnesia and behavioural abnormalities such as inertia and apathy can be demonstrated in these patients¹⁰. Thus they argued that, depending on the severity of the non-mnemonic deficits or the mode of presentation, cases of the Wernicke-Korsakoff syndrome may be misattributed as cases of alcoholic dementia. The neuropathologi-cal studies of Harper and colleagues^11–13 are in agreement with this view. They found that two thirds of the alcoholics coming to post-mortem in their unit had lesions of thiamine deficiency; of these, only a third had received a clinical diagnosis of the Wernicke-Korsakoff syndrome in life and, in the remainder, the most common diagnosis was dementia.

This evidence points to the conclusion that subcortical lesions caused by thiamine malnutrition can be sufficient to explain both amnesia and dementia witnessed in alcoholics. However, Victor and Adams⁹ and Torvik and colleagues⁸ did not suggest that allcases of alcoholic dementia are unrecognized cases of the Wernicke-Korsakoff syndrome. Both considered that superadded cerebral lesions may explain the dementia-like presentation of a proportion of patients with the Wernicke-Korsakoff syndrome. Because these additional lesions can be attributed to a variety of pathological processes, including chronic hepatocerebral degeneration, communi-cating hydrocephalus, Alzheimer’s disease and ischaemic infarction, they argued that there is no need to invoke a special process of alcohol neurotoxicity. Some studies have supported this hypothesis. For example, Kasahara and colleagues¹⁴ compared young and old alcoholics and found evidence of dementia only in the older group; most of these patients had additional medical diagnoses including hypertension, liver disease and cardiomyopathy and no case of dementia could be accounted for by the direct effect of alcohol alone.

Although these clinical studies suggest that dementia in alcoholics can be caused by factors other than the direct effect of alcohol on brain function, the neuropathological evidence is that alcohol is neu-rotoxic. The question therefore still remains as to whether this is sufficient to cause cortical damage and frank dementia. Carefully controlled neuropathological studies have frequently found whole brain atrophy involving grey and white matter in chronic alcoholics with and without evidence of thiamine depletion as indicated by the presence or absence of subcortical Wernicke-type lesions^15–18 . This atrophy appears to reflect reductions of white matter volume, neuronal loss in the dorsal frontal association cortex and neuronal shrinkage in other cortical areas such as the cingulate gyrus and the precentral motor cortex^19–23. The contribution of liver failure to this neuropathology has been assessed in several studies and the bulk of evidence suggests that cirrhosis alone does not account for the brain atrophy witnessed in alcoholics^16,17,24. Other findings suggest that thiamine depletion in the context of alcohol ingestion may con-tribute to this neuropathological picture even when this is insufficient to cause the subcortical lesions of the Wernicke-Korsakoff syndrome.

Thiamine is an essential co-enzyme for three enzyme complexes (pyruvate dehydrogenase, alpha-ketoglutarate dehydrogenase and transketolase) involved in mitochondrial oxidative phosphorylation and a number of cellular synthetic processes. Animal models have shown that lack of thiamine leads to mitochondrial dysfunction and cell death consequent upon oxidative stress, glutamatergic excitotoxicity and inflammatory responses; as these mechanisms are also active in other forms of neurodegenerative dementia, it is possible that thiamine depletion, as well as causing diencephalic vascular lesions, also directly causes cortical cell damage secondary to these actions²⁵. Animal models have also shown that alcohol alone can lead to cell damage via oxidative stress^26–28 and that cortical cell death is greater in the presence of both alcohol and thiamine depletion than either factor alone²⁹.

Thiamine deficiency is common in alcoholics due to a combi-nation of inadequate diet, poor absorption secondary to the effect of alcohol on the gastrointestinal tract and poor cellular utilization due to the direct action of alcohol on intracellular mechanisms^30,31. Thus it is likely that, even in alcoholics who have not developed Wernicke-Korsakoff syndrome, reduced thiamine availability will be present in a significant proportion and that this contributes to the development of dementia along with alcohol via a joint action on cortical neurones. Evidence that alcohol and thiamine deficiency are synergistic in man also comes from findings that the cortical patho-logical changes witnessed in alcoholics without evidence of thiamine depletion are similar to but less severe than those with co-existent subcortical Wernicke lesions^{16,18,21,24,32 –34} .

In summary, it can be concluded from these studies that: (i) alco-hol itself can cause neuronal damage including cell death; (ii) liver disease perseis not a major factor in the aetiology of the neu-ropathological cerebral changes seen in alcoholics; and (iii) thiamine malnutrition potentiates the neurotoxic effect of alcohol on the brain and it is likely that frank dementia in alcoholics is due to this com-bined action rather than the effect of alcohol alone.

What is the relationship between alcohol ingestion and the more common, milder cognitive impairment seen in alcoholics? The advent of invivoneuroimaging has enabled clinicopathological correla-tional studies to clarify the consequences of alcohol consumption on brain function and cognition in alcoholics uncomplicated by liver disease or thiamine depletion. Well-controlled, prospective studies using both CT and MRI have reliably confirmed the presence of cerebral shrinkage in such alcoholics in life involving both cortical grey and subcortical white matter^35–42, which is more pronounced in older patients⁴³ and more apparent in the frontal lobe^44,45. One early imaging study highlighted the possibility that women are more vul-nerable to the effects of alcohol in that brain atrophy was equivalent in age-matched men and women despite the women having shorter or less severe drinking histories⁴⁶. This remains controversial, how-ever, as some later studies^47,48 but not all^49,50 have replicated this finding.

The development of the MRI modality of diffusion tensor imaging (DTI) has allowed a more detailed assessment of the white matter vol-ume reductions found in postmortem and conventional MRI studies.

DTI depends on water diffusion being constrained by cell membranes or myelin sheaths and, when abnormal, may reflect pathology of myelin microstructure. Studies of alcoholics without complications of liver disease or a history of Wernicke’s encephalopathy have found that white matter abnormalities are widespread in life, being present in areas not previously detected by conventional MRI^51–53.DTI data can also be used to examine specific fibre pathways using the tech-nique of tractography. When applied to alcoholics, abnormalities in the frontal forceps, internal and external capsules, fornix, and supe-rior cingulate and longitudinal fasciculi have been detected, and the severity of these abnormalities was found to correlate with lifetime alcohol consumption⁵⁴. In the same study, when matched for alcohol exposure, alcoholic women showed more DTI signs of white matter degradation than alcoholic men in several fibre bundles, suggesting that any sex differences in the vulnerability to alcohol may be neural tissue specific⁵⁴.

An important observation is that the percentage of alcoholics with evidence of cerebral damage on brain scans is far in excess of that noted in neuropathological studies¹⁵. This discrepancy is probably explained by the finding that brain changes are reversible with continuing abstinence^{37,41,42,46,5563}

Only gold members can continue reading. Log In or Register to continue