Possible factors associated with epidemic neuropathy in Cuba

Abstract: This study included 115 patients with epidemic neuropathy being treated at the Center for Medical and Surgical Research (CIMEQ) in Havana from March 15 through April 30, 1993, and 114 controls randomly selected from the community. Using univariate analysis, significant associations were found, the most important being with smoking (odds ratio = 3.45 [95% confidence interval=1.85- 6.35] and with the consumption of uncertified edible fat (odds ratio = 2.77 [95% confidence interval= 1.34- 5.8]). Ratification was carried out using multivariate logistical regression analysis. Association (odds ratio = 8.8 [95% confidence interval= 2.58-30.55] with the presence of antibodies against the Coxsackie virus A-9 strain 47 was also found among the 182 individuals who had serological tests. The data suggests that toxic factors derived from smoking, plus deficiencies of some nutrients due to dietary changes, as well as contact with the isolated virus in one of the patients, facilitated the clinical expression of neuropathy and therefore the origin of the epidemic.

Subject headings:NEURITIS, OPTIC/epidemiology; POLYNEURITIS/epidemiology; RISK FACTORS; COXSACKIE VIRUSESA/isolation & purification; LOGISTIC REGRESSION; CUBA

In late 1991, an increase was observed in the incidence of optic neuritis among the adult population of Pinar del Río province, the westernmost region of Cuba. An even more significant increase was apparent during the first six months of 1992, and this situation extended to the rest of Cuba's provinces during 1993.

In the initial phase of the neuropathy epidemic, optic neuritis was the main form of presentation, exhibiting the following characteristics: diminished visual acuity; central or cecocentral scotoma; decrease or loss of bilateral and/or color vision; and insidious evolution within a period of 1-2 months. Later, cases began to appear of peripheral neuropathy, characterized by acroparesthesia, difficulty in walking and an increase in urinary frequency, among other disorders.

This process, known as epidemic neuropathy (EN), was also characterized by: susceptibility for both sexes, with the optic form predominant in males and the peripheral form predominant in females; significant risks for the 25-64 year age group; and heterogeneous geographical distribution.1-2

In general, toxic-nutritional and biological factors have been cited as the cause of such neuromyelopathies, although it should be noted that in some countries these processes have not been clarified entirely.3-4

The aforementioned circumstances clearly pointed to the need to perform studies to contribute to a better understanding of the factors associated with the appearance of the epidemic, which in turn would allow us to develop more effective preventive and control measures, and to establish a basis for new research.

Methods

The sample was composed of 115 adults suffering from EN living in Havana City, Havana and Pinar del Río provinces. Of these, 29 (25.2%) presented optic neuritis; 29 (25.2%) had peripheral polyneuropathy; and another 57 (49.6%) patients suffered from mixed forms. All were referred with presumed diagnoses for admission to the Center for Medical and Surgical Research (CIMEQ) in Havana City between March 15 and April 30, 1993, each with an evolution of less than 20 days. There, each case was reviewed and the diagnosis confirmed by the appropriate and previously trained specialists, based upon diagnostic criteria (Appendix 1) established nationwide.2,5

Simultaneously, 114 adult subjects not affected by EN, of similar ages (within 10 years) were randomly selected as the control group among relatives, friends and neighbors who voluntarily decided to participate and lived in the same area of the family doctor who attended each case.

The 229 individuals included in the study were examined by the appropriate CIMEQ specialists. Additionally, two specialists in Comprehensive General Medicine interviewed each participant, applying an epidemiological survey to obtain personal data, individual exposure to toxic or infectious elements and other factors that might be related to the process, as previously defined (Appendix 2).

A serological study was applied to 182 individuals (100 patients and 82 control subjects) consisting of a neutralization test against the Coxsackie virus A-9, strain 47, isolated in the Enterovirus Laboratory of the Pedro Kourí Institute of Tropical Medicine (IPK) from a sample of cephalorachidial liquid (CRL) taken from a patient suffering from EN6. A 1:10 or greater dilution was considered positive; and at the same time, as the criterion for determining infection or exposure to the biological agent.

Logarithmic transformation of antibody titers was made through a geometric mean calculation of titers for cases and controls, respectively, and the difference between them was estimated using the MICROSTAT software mean hypothesis test.

All the information collected was introduced in a database and processed using Epiinfo Software Version 5.1, first with a univariate analysis, and later with a multivariate analysis, using the RELODI multivariate logistical regression software package. The disparity ratio (DR) was considered as an association measure, and its confidence interval (CI) with a 95% reliability. Regression model adjustment was made step-by-step considering those associated variables which had a value >1, and with statistical significance those values of Z>1.96. Finally, a multivariate logistic regression analysis was carried out with the 182 subjects (100 cases and 82 controls), applying the serological study to assess viral agent association plus all other variables (excluding those with values of Z<1 during the previous logistic regression with all cases and controls) in the entity occurrence.

Results

The largest number of patients studied was in the 25-64 year age group, with neither sex predominating. Racially, there were more white patients, followed by black and mestizo. No significant differences were found in these variables between patients and the control group. (Table 1). Furthermore, no significant differences in occupational categories were identified, although the occupations most frequently found in both groups were service workers, technicians, professionals and military personnel.

Table 1 . Percentage of cases and controls according to age, sex and color of skin.

Age group	Cases n = 115 (%)	Controls n = 114 (%)	C2 Value of p
15 - 19	4.3	0.9
20 - 24	1.7	3.5
25 - 44	53.9	54.4
45 - 64	37.4	34.2
65 and over	2.7	7.0
Sex			0.6415
Male	50.4	53.5
Female	49.6	46.5
Skin color			0.6823
White	62.6	64.0
Black	25.2	21.1
Mestizo	12.2	14.9

 

When the univariate analysis of variables or risk factors was performed (Table 2), a significant association was determined only with smoking (DR = 3.45 [95%CI = 1.85 - 6.35]); consumption during the last three months of edible fats not distributed by the government, and therefore considered uncertified (DR = 2.77 [95% CI = 1.34 - 5.8]); regular consumption during at least three months of the last year of some medications (DR = 1.99 [95% CI = 1.13 - 3.51]); and the presence of vectors in the home, most frequently identified as flies and mosquitoes (DR = 2.53 [95% CI = 1.25 – 5.0]). A positive serology against the Coxsackie A-9 virus, strain 47 (IPK) was also significantly associated (DR = 4.61 [95% CI = 1.61 - 13.19]). Concerning the use of medications: no different drug or excessive use of any was found, which varied from the controls (Table 3).

Table 2. Univariate analysis of associated factors.

Factors	Cases n = 115 (No.)	Controls n = 114 (No.)	Disparity Ratio (95% CI)	Value of p
Smoking	85	54	3.45 (1.85-6.35)	0.00001
Pesticide	9	11	0.80 (0.29-2.19)	0.6251
Alcoholic beverage consumption	24	28	0.81 (0.41-1.58)	0.5049
Coffee consumption	89	75	1.78 (0.95-3.35)	0.0515
Infusion consumption	72	82	0.65 (0.36-1.19)	0.1328
Non-certified fat consumption	34	15	2.77 (1.34-5.80)	0.0024
Medicine consumption	65	45	1.99 (1.13-3.51)	0.0098
Previous diseases	27	16	1.86 (0.89-3.92)	0.0721
Vectors	96	76	2.53 (1.25-5.00)	0.0032
Domestic animals	48	50	0.92 (0.52-1.60)	0.7457
Previous contacts with EN	17	27	0.56 (0.27-1.16)	0.0873
Positive serology	95 (N=100)	66 (N=82)	4.61(1.61-13.19)	0.0022

Table 3. Cases and controls according to drug consumption.

Medication	Cases (No.)	Controls (No.)
Metronidazole	11	4
Indometazine	16	14
Meprobamate	7	3
Sulpha	4	3
Diazepam	6	5
PPG (ateromixol)	5	10
Acetyl Salicylic Acid	5	2

Previously existing illnesses (three months before the onset of symptoms for patients or three months before applying the survey for the control group) did not show any association, nor did previous contact with patients suffering from epidemic neuropathy (EN).

No significant differences between patients and controls were found concerning the consumption of infusions or herb teas from various plants (caña santa, linden, orange leaves, etc.).

Performing the analysis using the multivariate logistic regression for the most important variables or factors ( Table 4), we found once again an association with smoking (DR = 2.99 [95% CI = 1.57 - 5.69]); consumption of uncertified edible fats (DR = 3.30 [95% CI = 1.55 - 7.03]); and vectors in the home (DR = 2.82 [95% CI = 1.38 - 5.76]).

Table 4. Factors associated with multiple logistic regression.

Factor	Z Values	Disparity Ratio (95% CI)
Smoking	3.332727	2.99 (1.57 - 5.69)
Coffee consumption	1.000454	1.44 (0.70 - 2.92)
Infusion consumption	-2.030893	0.52 (0.28 - 0.98)
Medication consumption	1.789117	1.71 (0.94 - 3.10)
Non-certified fat consumption	3.096950	3.30 (1.55 - 7.03)
Previous diseases	0.722889	1.32 (0.62 - 2.82)
Contacts with EN patients	-0.957482	0.68 (0.32 - 1.48)
Vectors	2.833105	2.82 (1.38 - 5.76)

Table 5. Distribution of cases and controls according to neutralizing antibody study against the A-9 Coxsackie virus, strain 47.

Neutralizing antibody titers	CasesNo. %	ControlsNo. %	TotalNo. %
<10	5 5.0	16 19.6	21 11.6
10	14 14.0	8 9.7	22 12.1
20	27 27.0	18 21.9	45 24.7
40	20 20.0	14 17.1	34 18.7
80	13 13.0	14 17.1	27 14.8
160	10 10.0	8 9.7	18 9.9
320	11 11.0	4 4.9	15 8.2
Total	100 100.0	82 100.0	182 100.2

Table 5 shows the results of titers from antibody neutralizers for cases and controls against the A-9 Coxsackie virus strain 47 (IPK). The greater proportion of negative sera was found in the control group, and the greater proportion of high titers was detected in the patient group (>1:80). Furthermore, the geometric mean of the antibody titer was 35.9 among the patients and 20.8 among the control group, a statistically significant difference (T = 2.4878, p = 6.882E-03).

A strong association with positive serology was also observed (DR = 8.88[95% CI = 2.58 - 30.55]), and associations were maintained with smoking (DR = 2.73 [95% CI = 1.11 - 6.71]) and vector presence in the home (DR = 5.17 [95% CI = 2.37 - 11.3]) when logistical regression analysis was performed (Table 6).

Table 6. Factors associated with multiple logistic regression, including serological test.*

Factor	Z Values	Disparity Ratio (95% CI)
Smoking	2.12198	2.27 (1.06 - 4.87)
Coffee consumption	1.085135	1.59 (0.68 - 3.69)
Infusion consumption	-1.820544	0.49 (0.24 - 1.05)
Medication consumption	1.718865	1.82 (0.91 - 3.64)
Non-certified fat consumption	2.193603	2.73 (1.11 - 6.71)
Vectors	4.130548	5.17 (2.37 - 11.30)
Serology +	3.463935	8.88 (2.58 - 30.55)

* Cases = 100. Controls = 82.

Discussion

Tropical neuromyelopathy outbreaks have been described in the Caribbean for over 100 years7. In this particular outbreak, the dynamic of its appearance was at first that of an optic neuritis; some cases later accompanied by peripheral manifestations; and, finally, the appearance of some patients presenting only peripheral neuropathy. In our country, there is no record of any similar process of such magnitude.

While it may be true that this clinical manifestation may have toxic, deficiency-related, metabolic, or infectious origins--or a combination of these--identification of the fundamental or necessary cause is not always easy to determine, sometimes requiring lengthy research. Thus far, we have established some of the risk factors in this complex multi-origin phenomenon.

It appears that smoking has been an influential factor in this process, more so since reports have demonstrated that cyanide, nicotine and other elements produced by cigarette burning, accompanied by a vitamin B12 deficiency, are important toxic agents in smoking-induced amblyopia.2

The consumption of uncertified edible fats occurred when some patients bought lard from people who bred pigs and other animals, although in no case was this fat adulterated or illegally produced. This indirectly suggests that patients found the availability of certified fats deficient, and thus purchased uncertified fats in rather large quantities. We do not discard the possibility that other products consumed might have been toxic; or that these, consumed as substitutes for other products, might cause a nutritional deficiency. This is how we identified an increase in the consumption of infusions or herb teas as a new element in the normal diet, as a result of the special economic situation facing the country. Similarly, some authors9 have suggested that the sudden deficiency of Vitamin B complex, accompanied by deficiencies of certain essential amino acids and an increase in the consumption of cane sugar, gave rise to the epidemic.

It is important to recall, although this is not our case, that chronic cyanide intoxication as a consequence of excess consumption of cassava bread (casabe) and cyanogenic sausages has produced myeloneuropahy outbreaks, particularly in Africa.10,11

As for medication consumption: it was not possible to demonstrate that the use of new products or excessive use of medications was associated with this illness.

The lack of relation with previous diseases, particularly of infectious origin, does not permit us to support the hypothesis of a biological agent as the main cause, especially in view of the fact that there were few references to previous contacts with patients suffering from EN. Nevertheless, the participation of a biological agent cannot be totally ruled out, because of evidence of the isolation of four strains of Enterovirus belonging to the Coxsackie group. Additionally, in the CRL of other patients presenting the disease, a slight non-progressive cytopathic effect has been observed, which might be a possible viral agent not yet identified12. Another fact to underscore is that 30% of our patients reported fever at the onset of their symptoms, an element that further supports our hypothesis that a possible infectious agent may be associated with the outcome of the process under study.

It should be considered that many biological agents, like the Enterovirus, target several organs and have a broad clinical spectrum with the predominance of unapparent or subclinical infectious forms, which could explain the lack of reference to contact with other patients or the fact that some control subjects became ill without clinical manifestations of the disease.

In Japan in the sixties, a subacute myelo-optic neuropathy (SMON) occurred in which a possible infectious cause was also studied and in which some researchers isolated the type 21 Echo virus (Shingu et al.); others the Coxsackie A (Miyahara); and still others an agent with a slight cytopathic effect (Inoue et al.), none of which could be sustained because their observations were not confirmed in other laboratories. So the use of a drug (clioquinol) was established as the possible cause of that epidemic.13 It is important to emphasize that Inoue et al. detected a cytopathic effect in BAT-6 cells from samples of five patients in Okayama; eight patients in Osaka; and 23 cases in Hokkaido during the outbreak in Japan.14

We should insist upon the fact that determining the cause of optic neuritis has been and is frequently problematic, and thus we should recall the Benedict, Carroll and Otradovec series where several causes were put forward, including an infectious one.15

If an enterovirus, like the one we studied, were associated with this outbreak, this would also explain the association found with vectors, representing a possible indirect indicator of the hygienic and sanitary situation in the area where the patients live, or the expression of a mechanical transmission of the virus. In both cases, association with the Coxsackie A-9 virus, strain 47, should remain valid when analyzing the 100 cases and the 82 control subjects with serological studies, since this was a factor strongly associated with the disease, in addition to smoking, an important toxic component already mentioned. This evidence is more clearly reaffirmed upon identification of geometric means titers of neutralizing antibodies against the studied virus, which were significantly greater in patients (35.9) than in the control subjects (20.8).

It was not possible to perform pair sera studies to demonstrate the seroconversion, since the patients studied had approximately 20 days of evolution, due to the slow or subacute form of presentation.

The presence of infectious agents, associated even with chronic manifestations of the nervous system, is demonstrated by the fact that in four patients in Houston, three of them with multiple sclerosis and one with an amyotrophic lateral sclerosis, a virus was detected in the cephalo-rachidial liquid and patients also presented neutralizing antibodies against the isolated agent, whose antigenic characteristics showed a relation with the Inoue virus identified in Japan.16

Our serological studies were more encouraging than those of Kono et al. because when performing their neutralizing test studies with the Sato strain of the BAT-6 culture, the titers resulted negative in a dilution of 1:4 for patients and control subjects alike, the opposite of that found by Kimura et al., who identified positive antibody titers in a dilution of 1:5 to 1:10 in 13 of the 15 patients with SMON in Japan.17

It appears that the toxic factor derived from smoking, plus possible deficiencies in some nutrients caused by dietary changes, created conditions in the host, which during aggression of a biological agent--the isolated A-9 Coxsackie virus-- facilitated the clinical expression of this disease, and hence, the origin of this outbreak.

Our results, though not final, sustain a hypothesis for future research, and also permit the adoption of some immediate measures for prevention and control.

Appendix 1.

Criteria for defining the various clinical forms of epidemic neuropathy

1. Evaluation of peripheral symptoms to confirm diagnosis of the peripheral form.

Main criteria:

1.1 Peripheral sensitive symptoms of irritation, such as: pins-and-needles; cramps; numbness; hot flashes.

1.2 Sensibility disturbances, demonstrated by decreased perception of the 128-cycle diapason vibrations and of cotton and pin tracings, predominantly distal (fingers and toes).

1.3 Disturbances of osteo-tendinous reflexes of the lower limbs in a bilateral and generally symmetric form, with a decrease or absence of the Achillean reflexes and with or without patella hyperreflex.

Minor criteria :

1.4 Disturbances in the micturition process with an urgent desire to micturate; nicturia; pollakiuria; incontinence and polyuria.

1.5 Neurovegetative disturbances, such as cold or hot flashes; sweating of hands and feet; palpitations; tachycardia.

1.6 Others: hypoacusia; dysphagia; dysphonia; sensitive ataxia; constipation; diarrhea; sexual impotence; irritability; sleep disturbances.

The patient should exhibit three of the main diagnostic criteria or two main and at least one minor criteria (the 1.1 criterion should always be present).

2. Evaluation of visual symptoms for diagnosis of the optic form.

Main criteria:

2.1 Decrease of visual acuity (0.8 or less).

2.2 Disturbance of color vision: failure in two or more of the first eight strips of the Ishihara
      test, 1992 edition.

2.3 Absolute or relative bilateral (green or red) scotoma of the central type, within the 100 of
      central vision or the cecocentral type.

2.4 Disturbances in the contrast vision sensitivity test using the Vistech letter (less than 18).

2.5 Bilateral defect of the optic fiber layer (wedge in the maculopapillary bundle), which
      usually appears after one month of evolution.

Minor criteria:

2.6 Paleness in the temporal sector of the optic papilla in patients with over one month of
      evolution.

2.7 Photophobia, ocular burning sensation, lateral versions in the indented wheel.

Diagnosis is confirmed with the presence of at least four of the main criteria. The minor criteria are not necessary for diagnosis, but their presence reinforces its certainty.

3. Evaluation of patients with visual and peripheral symptoms.

The visual and peripheral symptoms should be evaluated separately, as indicated above. If the patient exhibits diagnostic requirements for the two clinical forms, the case shall be classified as a mixed form.

APPENDIX 2. Definitions of variables or risk factors

Variable	Definition
Occupation	Occupation during the last three months
Skin Color	Skin color of the subjects
Pesticide	Those subjects exposed to a pesticide during the last year
Smoking	Smoking at least five cigarettes a day for more than one month before the test
Alcoholic beverage consumption	Alcoholic beverage consumption at least three times a week during the last three months
Coffee consumption	Coffee consumption (at least one cup daily) for more than one month before the test
Consumption of infusions or herb teas	Consumption of infusions or herb teas (one glass daily) during the last three months
Medication consumption	Daily medication consumption for at least three months during the last year
Non-certified soap consumption	When the soap was acquired from other than official government sales
Non-certified fat consumption	When during the last three3 months the subjects acquired fat for their consumption from a non-certified source
Previous contacts with EN	Personal or direct contact with patients suffering from EN during the last three months
Animals	Presence of cats, dogs, birds and other domestic animals in the home
Vectors	Presence of mosquitoes, roaches, flies and rodents in the home
Previous diseases	Any infectious disease suffered during the previous three months
Vaccines	If the subject was vaccinated or injected with other substances or drugs during the last year
Previous travels	If the subjects traveled abroad during the last year
Positive serology	Titers greater than 1:10 of neutralizing antibodies

References

1. Ramírez A, Rodríguez R, Marrero A, Mesa G, Galindo MA, Iñiguez L. Neuropatía epidémica cubana. Breve reseña epidemiológica. Bol. Epidemiol. IPK 1993; (1): 1-4.
2. Cuba. Ministry of Public Health. International notes: Epidemic neurophathy. Cuba 1991-1994. MMWR 1994; 43(10): 183-92.
3. Schoenberg BS. Clinical neuroepidemiology in developing countries. Neuroepidemiology 1982;1:137-42.
4. Román GC, Spencer PS, Schoenberg BS. Tropical myeloneuropathies: the hidden endemias. Neurology 1985; 35: 1158-70.
5. Cuba. Ministerio de Salud Pública. Criterios de diagnóstico clínico de neuropatía epidémica. Bol Epidemiol IPK 1993; (1): 5-7.
6. Más Lago P, Rodríguez MP, Guzmán MG, Alvarez M, Mazio V, Ancheta O, et al. Resultados preliminares de laboratorio virológico en estudio de casos de neuropatía epidémica en Cuba. Bol Epidemiol IPK 1993; (1): 7-8.
7. Scott HH. An investigation into an acute outbreak of central neuritis. An Trop Med Parasitol 1918; 12:109-96.
8. Oku H, Fukushima K, Miyata M, Wakakura M, Ishikama S. Cyanide with vitamin B12 deficiency as the cause of experimental tobacco amblyopia. Acta Soc Ophthalmol Jan 1991;95(2):158-64.
9. Román GC. Epidemic neuropathy in Cuba: a plea to end the United States economic embargo on a humanitarian basis. Neurology 1994;44:1784-6.
10. Tylleskar T, Banea M, Bikangi N, Fresco L, Persson L.A, Rosling H. Epidemiological evidence From Zaire for a dietary etiology of Konzo, an upper motor neuron disease. Bull World Health Organ 1991; 69:581-9.
11. Mozambique Ministry of Health. Mantakassa: an epidemic of spastic paraparesis associated with Chronic cyanide intoxication in a cassava staple area of Mozambique. 1. Epidemiology and clinical and laboratory findings in patients. Bull World Health Organ 1984;62:477-84.
12. Más Lago P, Pelegrino JL, Guzmán M, Capó V, Rodríguez L, Rodríguez P, et al. Neuropatía epidémica cubana. Parte I. Aislamiento viral. Rev Cubana Med Trop 1995;47(1):11-5.
13. Kono R. Introductory review of subacute myelo-optico-neuropathy (SMOM) and its studies done by the SOMN research commission. Jpn J Med Science Biol 1975;28:1-21.
14. Kono R., Keizo Y. History of etiological studies. So-called Inoue virus. In: History and review of SMON research. Tokyo: Japanese Ministry of Health and Welfare Research Grant, 1989; 33-9.
15. Hayreh SS. Neuritis óptica. Etiología. In: Tratado de neuro-oftalmología. Barcelona: Jims, 1985:415-31.
16. Melnick J.L., Seidel E, Inoue YK, Nishibe Y. Isolation of virus from the spinal fluid of three patients multiple sclerosis and one with amyotrophic lateral sclerosis. Lancet 1982; 1(10):830-3.
17. Tagaya I. Summarized report of research works by the members of the virology subsection. Microbiology section. Jpn J Med Sci Biol 1975; 28:197-202.

This article originally appeared in Spanish in the Revista Cubana de Medicina Tropical, Vol. 50, No. 1, (pp. 54-60), 1998. 

Copyright (c) 1999, MEDICC - Medical Education Cooperation with Cuba