Peripheral neuropathy in chronic alcoholism: a retrospective cross-sectional study in 76 subjects

A. Ammendola*,2, M. R. Tata2, C. Aurilio1,2, G. Ciccone2, D. Gemini1,2, E. Ammendola2, G. Ugolini2 and F. Argenzio1,2

Departments of Neurophysiopathology and
1 Clinical Toxicology, Faculty of Medicine — II University of Naples, Italy

Received 27 July 2000; in revised form 5 December 2000; accepted 19 January 2001


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
A consecutive sample of 76 chronic alcoholic patients was studied clinically, biochemically and electrophysiologically to assess clinical and/or subclinical signs of alcohol-related neuropathy as well as the most important and disputed risk factors for neuropathy such as age, parental history of alcoholism, nutritional status, alcoholic disease duration and total lifetime dose of ethanol (TLDE). The results show that alcohol-related neuropathy, especially when subclinical, seems to be frequent and mostly characterized by axonal degeneration of peripheral nerve fibres with earlier and more frequent involvement of sensory fibres and lower limbs. Moreover, positive family history of alcoholism, but above all alcoholic disease duration and TLDE, could be more important factors than malnutrition in determining neuropathy.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Peripheral neuropathy is a pathology which can be associated with chronic alcoholism. Although neuropathy is frequent among chronic alcoholics, there are no uniform data about its prevalence. The literature reports a 9–50% prevalence (Bolton, 1987Go; Charness et al., 1989Go; Victor, 1994Go; Monforte et al., 1995Go; Pessione et al., 1995Go; Vuadens and Bogouslasky, 1998).

Moreover, alcohol-related neuropathy is associated with several risk factors, such as malnutrition, thiamine deficiency, direct toxicity of alcohol and recently family history of alcoholism (Victor, 1975Go; Bosch et al., 1979Go; Claus et al., 1985Go; Huas et al., 1993Go; Dyck et al., 1994Go; Monforte et al., 1995Go; Pessione et al., 1995Go), but which of these plays a primary role in inducing neuropathy is still unclear (Estruch et al., 1993Go; Palliyath and Schwartz, 1993Go; Windebank, 1993Go).

In the present work, we carried out a retrospective, cross-sectional study on a group of chronic alcoholics to evaluate parameters in alcohol-related neuropathy, such as (1) prevalence of clinical and electrophysiological alterations, (2) type of electrophysiological alterations, (3) presence of correlations between some clinical parameters such as parental history of alcoholism, age, nutritional status, alcoholic disease duration and total lifetime dose of ethanol (TLDE), and peripheral neuropathy.


    PATIENTS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Patients
Seventy-six chronic alcoholics (55 men and 21 women), age range 24–69 years, were consecutively enrolled in this study over a period of 28 months. Each patient asked to be admitted to the hospital for alcohol abuse. All patients reported a consumption of >=100 g of alcohol per day over the 2 years preceding admission. We excluded all patients with other diseases which could damage the peripheral nervous system. To evaluate the real dose of alcohol consumed, a detailed history was taken and then confirmed by family members. Special events in each subject's life were used to jog their memory, such as military service, marriage, a new job, etc. (‘time look follow-back method’) (Sobell et al., 1979Go).

Methods
Each patient received assessment for the following aspects. (1) Careful history, paying particular attention to family history of alcoholism, defined as positive when father and/or mother presented alcohol abuse. (2) Clinical and neurological examination: neurological assessment was carried out using a neuropathy symptom score and neurological disability score (Dyck, 1988Go). Clinical peripheral neuropathy was considered when the patient had two or more of the following clinical abnormalities: muscle weakness, paraesthesia, symmetrically depressed or absent tendon reflexes and sensory deficit. (3) Calculation of TLDE expressed in kg of ethanol/kg of body weight; this was estimated by first multiplying the daily consumption of ethanol by the number of days in the periods exposed to alcohol and then dividing the product by the body weight of the patient when first admitted. Alcoholic disease duration was identified as the number of years corresponding to the sum of alcohol intake periods. (4) Laboratory tests: these included glucose, creatinine, cholesterol, triglycerides, electrolytes, serum aspartate and alanine aminotransferases, {gamma}-glutamyl transpeptidase, protein electrophoresis, lactate dehydrogenase, creatine kinase, aldolase, red blood cell, white blood cell and platelet counts, haematocrit, total protein, prealbumin, albumin, iron, transferrin, prothrombin time and ethanol levels in blood and urine. (5) Evaluation of nutritional status: comparison between the actual weight and the ideal weight; study of the lean body mass calculated from the circumference of upper non-dominant arm and the thickness of the tricipital skin fold (expression of total body fat); study of nutritional proteins, evaluating total protein, albumin, prealbumin, total lymphocytes, transferrin, and of serum folate with vitamin B1 by chromatographic method (pathological values <16 ng/ml) and vitamin B12 studied with a fluorimetric enzyme-linked assay (pathological values <=226 pg/ml, that is the 5th percentile of our general population). Caloric or protein malnutrition was considered whenever body weight was <90% of the ideal weight or a lean body mass was <90% of the normal value or at least three of the examined nutritional serum parameters were deranged (Durnin and Womersley, 1974Go; Blackburn et al., 1977Go; Burrit and Anderson, 1984Go). (6) Electroneurographic evaluation of the ulnar, median, peroneal and sural nerves: this involved a study of maximum motor conduction velocity of the median and peroneal nerves with their compound action potential amplitude; study of sensory orthodromic conduction velocity of the ulnar and sural nerves with their sensory-evoked potential (SEP) amplitude. The evaluated parameters were considered pathological when they differed from the average value ± 2.0 SD, obtained from a control group of 40 subjects aged between 30 and 60 years, who did not consume alcoholic drinks. Subclinical peripheral neuropathy was considered to be present if the patient had one or more of the above-reported electrophysiological parameters altered at least in two of the examined nerves, but no clinical symptoms or signs.

Data analysis
All 76 patients were subdivided into two groups: the first group was made up of patients with neuropathy, the second group were subjects without neuropathy. In these two groups, some clinical risk factors such as age, nutritional status, family history of alcoholism, alcoholic disease duration and TLDE were compared. Moreover, considering sural nerve SEP amplitude as a suitable parameter for the evaluation of sensory axonal dysfunction, a comparison and statistical evaluation (Student's t-test) of sural nerve SEP amplitude was performed between alcoholics ‘in toto’ and the control group and then among neuropathic alcoholics ‘in toto’ and neuropathic alcoholics respectively with advanced age, altered nutritional status, positive family history of alcoholism, long alcoholic disease duration and high TLDE. Advanced age, long alcoholic disease duration and high TLDE were considered when these parameters exceeded their respective average values obtained from our series of cases. Finally, risk factors for neuropathy (age, nutritional status, positive family history of alcoholism, alcoholic disease duration and TLDE) were correlated with sural nerve SEP amplitude in chronic alcoholics with neuropathy through the Pearson's correlation coefficient multivariate analysis.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Table 1Go reports the main clinical–instrumental findings from the study. The group of alcoholic patients was composed of 76 subjects, 55 men and 21 women (age range: 24–69 years; mean 43.2). The average alcoholic disease duration was 14.5 (minimum of 2; maximum of 40) years; the mean daily alcohol intake of patients was 284.0 (minimum of 100; maximum of 750) g; the average TLDE was 23.6 (minimum of 1.2; maximum of 147.5) kg of ethanol/kg of body weight; the nutritional status was altered in 13 (17.1%) cases; all with caloric malnutrition, 10 with protein malnutrition. With regard to vitamins, mean values (± SD) of both B1 and B12 were normal (vitamin B1 = 31.0 ± 11.8 ng/ml; vitamin B12 = 625 ± 423 pg/ml) and only three patients had thiamine values slightly lower than normal. Parental history of alcohol abuse was present in 35 patients (46%). Finally, alcoholic neuropathy was present in 51 subjects (67.1%), of whom 28 (36.8%) had only subclinical neuropathy. The patients with clinical neuropathy were all aware of the symptoms presented, even if the symptomatology was not disabling in most of them.


View this table:
[in this window]
[in a new window]
 
Table 1. Clinical and instrumental data of chronic alcoholics studied
 
Table 2Go reports the influence of risk factors for peripheral neuropathy (sex, age, nutritional status, family history of alcoholism, alcoholic disease duration and TLDE) in patients with and those without peripheral neuropathy. The clinical signs of neuropathy were essentially represented by hyposthenia, symmetrically depressed or absent deep tendon reflexes and sensory deficit mostly in the lower limbs. The mean age was higher in the neuropathic patients (45.3 ± 9.4 years) compared to patients without neuropathy (39.1 ± 7.7 years). As for gender, 37 of 55 men and 14 of 21 women had neuropathy, whereas 18 of 55 men and seven of 21 women had no signs of disease. Nutritional status in the neuropathic group was normal in 40 cases, but altered in 11 patients, whereas the group without neuropathy was characterized by normal nutritional status in 23 subjects and altered status in only two patients. Moreover, parental history of alcoholism was absent in 25 but present in 26 of the neuropathic patients, whereas the corresponding numbers were 16 and nine of the group without neuropathy. The mean alcoholic disease duration was significantly greater (P < 0.05) in the neuropathic alcoholics (16.2 ± 9.4 years) than in alcoholics without neuropathy (11.1 ± 8.2 years) and the mean TLDE was significantly greater (P < 0.05) in the neuropathic cases (27.9 ± 24.4 kg of ethanol/kg of body weight) compared to patients without neuropathy (14.8 ± 15.9 kg of ethanol/kg of body weight).


View this table:
[in this window]
[in a new window]
 
Table 2. Correlation among presence or absence of peripheral neuropathy and risk factors for neuropathy considered (sex, age, nutritional status, family history of alcoholism, alcoholic disease duration, mean total lifetime dose of ethanol)
 
Figure 1Go shows frequency, type and distribution of electroneurographic alterations (conduction velocity and evoked potential amplitude) in the 76 alcoholic patients. The electroneurographic alterations consisted fundamentally of slightly decreased sensory and/or motor conduction velocities, of a reduced SEP amplitude or its absence, and of a low composed muscle action potential amplitude. Furthermore, the most frequent changes concerned the sensory nerves and they predominated in the lower limb; the most altered parameter was SEP amplitude (lower limb = 51.3% and upper limb = 37.0% of the cases), followed in decreasing order by sensory conduction velocity (lower limb = 47.3% and upper limb = 37.0% of the subjects), motor-evoked potential amplitude (lower limb = 23.6% and upper limb = 7.4% of the patients) and motor conduction velocity (lower limb = 7.8% and upper limb = 3.7% of the cases).



View larger version (37K):
[in this window]
[in a new window]
 
Fig. 1. Frequency of kind of electroneurographic alterations in 51 patients with peripheral neuropathy. MCV, motor conduction velocity; MEP-AMP, motor-evoked potential amplitude; SCV, sensory conduction velocity; SEP-AMP, sensory-evoked potential amplitude. {square}, upper limb; {blacksquare}, lower limb.

 
Table 3Go reports the comparison and significance of the main electroneurographic data (evoked potential amplitude and nerve conduction velocity) between chronic alcoholics ‘in toto and controls. The alcoholics showed evoked potential amplitude and conduction velocity of all examined nerves lower than the controls and this decrease was statistically significant for median (motor conduction velocity: P < 0.01), peroneal (motor-evoked potential amplitude: P < 0.001; motor conduction velocity: P < 0.01), ulnar (sensory-evoked potential amplitude and sensory conduction velocity: P < 0.001) and sural (sensory-evoked potential amplitude and sensory conduction velocity: P < 0.001) nerves.


View this table:
[in this window]
[in a new window]
 
Table 3. Comparison and significance of the nerve conduction studies between controls and chronic alcoholics
 
Table 4Go shows the comparison and statistical significance of the mean sural nerve SEP amplitude between neuropathic alcoholics in toto’ and neuropathic alcoholics respectively with advanced age, altered nutritional status, positive family history of alcoholism, long alcoholic disease duration and high TLDE. The mean (± SD) SEP amplitude, compared to neuropathic alcoholics ‘in toto’ (2.5 ± 1.6 µV), more or less overlapped in neuropathic alcoholics with advanced age (2.6 ± 1.5 µV) and positive parental history of alcoholism (2.7 ± 1.7 µV), decreased, but not significantly, in neuropathic alcoholics with malnutrition (1.9 ± 1.7 µV) and with long alcoholic disease duration (2.3 ± 1.6 µV), but significantly reduced (P < 0.05) in neuropathic alcoholics with high TLDE (1.7 ± 1.3 µV).


View this table:
[in this window]
[in a new window]
 
Table 4. Comparison and significance of sural nerve sensory-evoked potential (SEP) mean amplitude among chronic alcoholics with peripheral neuropathy ‘in toto’ and neuropathic alcoholics respectively with advanced age, altered nutritional status, positive family history of alcoholism, long alcoholic disease duration and high total lifetime dose of ethanol (TLDE) consumed
 
In Table 5Go, we evaluated the relationship between sural nerve SEP amplitude and the risk factors for neuropathy considered (age, nutritional status, positive family history of alcoholism, alcoholic disease duration and TLDE) in chronic alcoholics with peripheral neuropathy. Pearson's correlation indicated a significant inverse correlation (P < 0.01) only between sural nerve SEP amplitude and both alcoholic disease duration and TLDE in these patients.


View this table:
[in this window]
[in a new window]
 
Table 5. Pearson's correlation of sural nerve sensory-evoked potential amplitude and risk factors for peripheral neuropathy in neuropathic chronic alcoholics (n = 51)
 

    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This retrospective study performed on 76 chronic alcoholics revealed frequent peripheral neuropathy, especially when clinical examination was followed by an electrophysiological investigation. In our series of cases, the prevalence of clinical neuropathy was 30.2%, but increased to 67.1% when patients with established clinical neuropathy were added to those with subclinical neuropathy.

Electroneurographic studies of alcoholic neuropathy show damaged sensory and motor nerve peripheral fibres caused by an axonal degeneration and consisting mainly of decreased sensory- and/or motor-evoked potential amplitudes with light involvement of conduction velocity (Kimura, 1989Go; Vuadens and Bogouslasky, 1998). Our electroneurographic findings indicate that the most frequent and earliest parameter to be affected is sensory-evoked potential amplitude, followed by sensory conduction velocity, motor-evoked potential amplitude and lastly motor conduction velocity. These results confirm the presence of mostly axonal neuropathy with more frequent and early involvement of sensory nervous fibres specially of the lower limbs.

As for aetiology and pathogenesis, the literature points out that alcoholic neuropathy is related to several factors: malnutrition, thiamine deficiency, direct toxicity of alcohol and, more recently, family history of alcoholism. According to some authors, alcoholic peripheral neuropathy is due mainly to malnutrition (Windebank, 1993Go), whereas according to other authors, neuropathy is related to the direct toxicity of alcohol on the peripheral nervous system (Estruch et al., 1993Go; Palliyath and Schwartz, 1993Go). In support of this latter hypothesis, an experimental investigation on animals showed that cytochrome P 4502E1, an ethanol-inducible isoenzyme of the P450-dependent pathway for ethanol oxidation in hepatocytes and neurons, may be involved in alcohol-related neurotoxicity (Wohrle et al., 1998Go). Moreover, a recent clinical cross-sectional study postulated that alcohol may have a dose-related toxic effect and could be considered an important risk factor for peripheral neuropathy (Monforte et al., 1995Go). With regard to family history of alcoholism and alcohol-related diseases, few studies have reported hereditary factors in chronic alcoholic patients in relation to hepatic (Hrubec and Omenn, 1981) or central nervous system diseases (Begleiter et al., 1983Go; Hill et al., 1986Go; Polich et al., 1994Go).

Our study indicates that, in chronic alcoholics with peripheral neuropathy, there are few cases with malnutrition (less than a quarter), a moderate prevalence of patients with positive family history of alcoholism (more than a half) and a statistically significant correlation between peripheral neuropathy and high TLDE. In agreement with Monforte et al. (1995), these findings suggest that alcohol could be an important risk factor for alcoholic neuropathy. On this basis, we cannot rule out positive family history of alcoholism as an important factor.

Furthermore, the hypothesis of a direct toxicity of alcohol on peripheral nervous fibres is strengthened by the fact that sural nerve SEP amplitude, a suitable parameter to evaluate sensory axonal dysfunction and the most frequently altered among the electroneurographic parameters in our series of cases, is significantly more decreased in neuropathic patients with high TLDE, than in subjects with altered nutritional status or positive family history of alcoholism.

Finally, a multivariate analysis performed to evaluate the independent effect of risk factors for neuropathy in relation to sural nerve SEP amplitude in neuropathic chronic alcoholics seems to support this hypothesis; based on Pearson's correlation coefficient study, SEP amplitude was significantly inversely correlated only to alcoholic disease duration and to TLDE, confirming a dose-dependent relationship between SEP amplitude decrease and increased alcohol consumption. As reported in our previous study (Ammendola et al., 2000Go) there seems to be a greater female sensitivity to the toxic effects of alcohol on peripheral nerve fibres.

In conclusion, our study indicates that in chronic alcoholism: (1) peripheral neuropathy seems to be very frequent especially when subclinical. Thus a routine use of electrophysiological investigations for chronic alcoholics could be helpful, especially for those patients receiving disulfiram treatment which can occasionally lead to neuropathy; (2) neurogenic damage is mostly characterized by axonal degeneration of sensory and motor nerve fibres with earlier and more frequent involvement of sensory nerves and lower limbs; (3) positive family history of alcoholism, but especially all alcoholic disease duration and TLDE, appear to be more important factors than malnutrition in determining neuropathy.


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
* Author to whom correspondence should be addressed at: Cattedra e Servizio di Neurofisiopatologia, Policlinico, Piazza L. Miraglia, 2, 80138 Napoli, Italy. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ammendola, A., Gemini, D., Iannaccone, S., Argenzio, F., Ciccone, G., Ammendola, E., Serio, L., Ugolini, G. and Bravaccio, F. (2000) Gender and peripheral neuropathy in chronic alcoholism: a clinical–electroneurographic study. Alcohol and Alcoholism 35, 368–371.[Abstract/Free Full Text]

Begleiter, H., Porjesz, B., Bihari, B. and Kissin, B. (1983) Event-related brain potentials in boys at risk for alcoholism. Science 225, 1493–1495.[ISI]

Blackburn, G. C., Bistrian, B. R., Maini, B. S., Schlamm, H. T. and Smith, M. F. (1977) Nutritional and metabolic assessment of the hospitalized patients. Journal of Parenteral and Enteral Nutrition 1, 11–22.[Medline]

Bolton, C. F. (1987) Metabolic neuropathy. In Clinical Electromyography, Brown, W. F. and Bolton, C. F. eds, pp. 245–281. Butterworth, Boston.

Bosch, E. P., Pelham, R. W., Rasool, C. G., Chattorjee, A., Lasch, R. W. and Brown, L. (1979) Animal models of alcoholic neuropathy: morphological, electrophysiological and biochemical findings. Muscle and Nerve 2, 133–144.[ISI][Medline]

Burrit, M. F. and Anderson, C. F. (1984) Laboratory assessment of nutritional status. Human Pathology 15, 130–133.[ISI][Medline]

Charness, M. E., Simon, R. P. and Greenberg, D. A. (1989) Ethanol and the nervous system. New England Journal of Medicine 17, 442–454.

Claus, D., Egger, R., Engelhardt, A., Neundorfer, B. and Warecka, K. (1985) Ethanol and polyneuropathy. Acta Neurologica Scandinavica 72, 312–316.[ISI][Medline]

Durnin, J. V. and Womersley, J. (1974) Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. British Journal of Nutrition 32, 77–97.[ISI][Medline]

Dyck, P. J. (1988) Detection, characterization and staging of polyneuropathy assessed in diabetics. Muscle and Nerve 11, 21–32.[ISI][Medline]

Dyck, P. J., Low, P. A. and Stevens, J. C. (1994) Diseases of peripheral nerves. In Clinical Neurology, Vol. 4, Joynt, J. R. ed., revised edition, chap. 51, pp. 1–126. Lippincott, Philadelphia.

Estruch, R., Nicolas, J. M., Villegas, E., Junqué, A. and Urbano-Marquèz, A. (1993) Relationship between ethanol-related diseases and nutritional status in chronically alcoholic men. Alcohol and Alcoholism 28, 543–550.[Abstract]

Hill, S. Y., Armstrong, J., Steinhauer, S. R., Baughman, T. and Zubon, J. (1986) Static ataxia as a psychobiological marker for alcoholism. Alcoholism: Clinical and Experimental Research 11, 345–348.[ISI]

Hrubech, Z. and Omenn, G. S. (1981) Evidence of genetic predisposition to alcoholic cirrhosis and psychosis: twin concordances for alcoholism and its biological end point by zygoty among male veterans. Alcoholism: Clinical and Experimental Research 5, 207–215.[ISI][Medline]

Huas, D., Allemand, H., Loiseau, D., Pessione, F. and Rueff, B. (1993) Prévalence du risque et des maladies liés à l'alcool dans la clientèle adulte du généraliste. Revue du Praticien: Medecine General 7, 39–44.

Kimura, J. (1989) Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 2nd edn. F. A. Davis, Philadelphia.

Monforte, R., Estruch, R., Valls-Solé, J., Nicolás, J., Villalta, J. and Urbano-Marquez, A. (1995) Autonomic and peripheral neuropathies in patients with chronic alcoholism. A dose-related toxic effect of alcohol. Archives of Neurology 52, 45–51.[Abstract]

Palliyath, S. and Schwartz, B. D. (1993) Peripheral nerve functions improve in chronic alcoholic patients on abstinence. Journal of Studies on Alcohol 54, 684–686.[ISI][Medline]

Pessione, F., Gerchstein, J. L. and Rueff, B. (1995) Parental history of alcoholism: a risk factor for alcohol-related peripheral neuropathies. Alcohol and Alcoholism 30, 749–754.[Abstract]

Polich, J., Pollock, V. E. and Bloom, F. E. (1994) Meta-analysis of P300 amplitude from males at risk for alcoholism. Psychological Bulletin 115, 55–73.[ISI][Medline]

Sobell, L. C., Maisto, S. A., Sobell, M. B. and Cooper, A. M. (1979) Reliability of alcohol abusers' self-reports of drinking behaviour. Behaviour Research Therapy 17, 157–160.[ISI][Medline]

Victor, M. (1975) Polyneuropathy due to nutritional deficiency and alcoholism. In Peripheral Neuropathy, Dyck, P. J., Thomas, P. K. and Lambert, E. H. eds, pp. 1030–1066. W. B. Saunders, Philadelphia.

Victor, M. (1994) Neurologic disorders due to alcoholism and malnutrition. In Clinical Neurology, Vol. 4, Joynt, J. R. ed., revised edition, chap. 61, pp. 1–94. Lippincott, Philadelphia.

Vuadens, P. and Bogouslavsky, J. (1998) Complications neurologiques liées à l'alcool. Encyclopédie Médico-Chirurgicale, Neurologie, 17-161–B-10, 8 p. Elsevier, Paris.

Windebank, J. A. (1993) Polyneuropathy due to nutritional deficiency and alcoholism. In Peripheral Neuropathy, 3rd edn, Vol. 2, Dyck, P. J. and Thomas, P. K. eds, pp. 1310–1321. W. B. Saunders, Philadelphia.

Wohrle, J. C., Spengos, K., Steinke, W., Goebel, H. H. and Hennerici, M. (1998) Alcohol-related acute axonal polyneuropathy: a differential diagnosis of Guillain–Barre syndrome. Archives of Neurology 55, 1329–1334.[Abstract/Free Full Text]