Favism and Hemolytic Anemia in Glucose-6-Phosphate Dehydrogenase-Deficient Subjects in North Sardinia
Tullio Meloni a , Gavino Forteleoni a , Angelo Dore a , Stefano Cutillo b a
Department of Pediatrics, University of Sassari; b
Department of Pediatrics, Ist Faculty of Medicina, Naples, Italy
Drugs, Favism, G-6-PD deficiency, Hemolytic anemia, Viral infections
The present paper reports the incidence from 1965 to 1979 of acute hemolytic anemia for a total of 948 cases in G-6-PD deficient subjects due to the ingestion of fresh or dried fava beans or certain drugs and to viral infections. The highest percentage of hemolytic crises was due to fresh fava beans (94.4%). No cases of favism were observed in breast-fed babies whose mothers had eaten fava beans or from pollen inhalation. The male sex proved to be the hardest hit. Hemoglobin values were lower than or equal to 7 g/dl in about 75% of males and 50% of females. Total bilirubin values were lower than 03 umol/l (6 mg/dl) in about 75% of males and 85% of females. Both the hemoglobin and bilirubin values were statistically significant. Mean trnsaminase values (SGPT) were significantly higher than those of normal controls. No correlation between favism and blood groups was found.
Glucose-6-phosphate dehydrogenase (G-6-PD) deficiency is an X-linked disorder, which can lead to acute hemolytic anemia following ingestion of fava beans and certain drugs and bacterial or viral infection.
Although many G-6-PD variants have been described, susceptibility to fava beans and some drugs has been mainly confined to the G-6-PD Mediterranean variant largely found in Sardinia. Its incidence in the North is different from that of the South. Salvidio et al.  reported an incidence of about 25% in the region around Cagliari and a frequency of 9% in the region of Sassari.
The present study reports the cases of acute hemolytic anemia following ingestion of fava beans and certain drugs or during infectious diseases observed at the Children's Hospital of Sassari from 1965 to 1979.
Subjects and Methods
This study includes 948 children aged 2-12 years admitted to the clinic for acute hemolytic anemia (Table I).
Table I. Incidence of hemolytic anemia from various agents in G-6-PD-deficient children from 1965 to 1979
| ||Number of children ||Females ||Males |
|n ||% ||n ||% |
|Fresh fava beans ||895 ||257 ||28.7 ||638 ||71.3 |
|Dried fava beans ||28 ||5 ||17.8 ||23 ||82.2 |
|Drugs ||24 ||6 ||25.0 ||18 ||75.0 |
|Viral hepatitis ||1 ||- ||- ||1 ||- |
|Total ||948 ||268 ||28.3 ||680 ||71.7 |
Diagnosis of favism or acute hemolytic anemia due to drugs or infectious disease was made on the basis of clinical symptoms, such as anemia, hemoglobinuria and incterus and confrmed by a screening test for G-6-PD deficiency . Blood groups were determined in all subjects immediately before transfusion and hemoglobin, reticulocyte count and bilirubin were evaluated in more than 70% of all subjects. Moreover, in the last 3 years serum glutamic-oxalacetic transaminase (SGOT), serum glutamic-pyruvic transaminase (SGPT) and HbA 2 , were determined in subjects with favism. Enzyme activity and red cell mosaicism for G-6-PD were evaluated 4 months after the hemolytic crisis in 41 females only.
Hemoglobin, hematocrit, bilirubin and transaminase (SGOT and SGPT) were determined by standard methods. HbA 2 levels were measured according to Huisman et al. : screening for G-6-PD was made by the Brewer's test  and the activity was evaluated spectrophotometically using the method of Bauisruzzi et al. . The degree of red cell G-6-PD mosaicism in females was established by the method of Sansone et al. .
Children are more susceptible to favism than adults [6-8]. A comparison of the number of favic subjects admitted to the pediatric ward with that of adult wards in Sassari from 1975 to 1979 showed that children constituted 70.7% of all admissions (fig. 1).
The male sex is obviously the hardest hit by hemolytic anemia from fresh or dried fava beans, drugs or infection (Table I). The highest incidence of favism or drug-induced hemolytic anemia occurred in children aged between 2 and 6 years (fig. 2: males 80.7% and females 70.3%). Children below 2 years of age made up only 4.5% (males 5.3% and females 2.3%), whereas those over 6 constituted 22.3 % (males 19.3%and females 29.9%).
Regarding the period of the year, the hemolytic crises occurred largely during the months of May and June with sporadic cases in April. Fresh fava beans were the major cause of hemolytic anemia (895 cases), while dried beans were the cause in only 28 cases. The latter were observed mainly during the months of January and February. In the last few years, 3 cases of hemolytic anemia after ingestion of frozen fava beans were seen. No cases of hemolysis in breast-fed babies whose mothers had eaten fava beans were observed.
Drug-induced hemolytic anemia was observed in only 24 cases, in spite of the fact that antipyretic drugs are widely used in pediatrics. No differences with regard to age and sex were observed. The anemia was found to be less serious, blood transfusions not being necessary in about half of the cases.
Antipyretic drugs associated with sulphonamides were most frequently involved although we observed 1 case of anemia from aspirin alone, 2 from chloramphenicol and 2 from naphthalene. Anemia caused by chloramphenicol was observed in 2 G-6-PD deficiency children with typhoid fever.
Severe hemolytic anemia was observed in a G-6-PD-deficient boy with viral hepatitis who had been admitted for favism 2 years before. Moderate anemia was also found in one normal and four other G-6PD-deficient children with viral hepatitis.
Clinical symptoms, i.e. pallor, mild icterus and hemoglobinuria were usually observed 24-48 h after ingestion of fava beans, whether fresh or dried. Hemolytic anemia exceptionally occurred later. The same was not true of drug-induced hemolytic anemia, since a longer time elapsed between the ingestion of drugs and the appearance of clinical symptoms (usually 4-5 days). Hemoglobin values proved to be lower than 7 g/dl in 74.15% of males and in 53.45% of females (fig. 3). Total bilirubin levels were elevated in all cases and did exceed 103 ,µmol/l (6 mg/dl) in 23.85% of males and in 16.08% of females (fig. 4).
Mean transaminase values (SGOT 9.9 ± 8.4, SGPT 31.7 ± 21.7) were significantly higher (p < 0.001 ) than those of normlal controls (Table II). Of 41 females evaluated 4 months after a hemolytic episode, G6-PD activity was undetectable in 8 (homozygotes) while the remaining 33 subjects (heterozygotes) had a mean activity of 2.6 ± 1.3 IU (normal values 6.5 ± 1.8). These results are statistically significant (p < 0.05) when compared with those of 32 heterozygous females who never experienced favism (mean activity 3.3 ± 1.1 IU). No significant correlation ( c 2 = 2.5818; p = 0.4607) between blood groups and favism was oberved (Table III).
Table II. Mean values (±SD) of serum GOT and GPT in G-6-PD-deficient children with favism and in normal controls
| ||Number of cases ||SGOT ||SGPT |
|Children ||112 ||9.9±8.4 + ||31.7±21.7* |
|Controls ||40 ||8.1±3.5 + ||8.6±3.9* |
*=p < 0.001.
Favism is a limited aspect of G-6-PD deficiency and is confined to relatively small geographical areas such as the Mediterranean region, the Far East and Southern Asia, although sporadic cases have also been described elsewhere [6, 7]. The highest incidence, about 15% of all G-6-PD-deficient subjects, has been described in Sardinia . The G-6-PD enzyme of subjects sensitive to fava beans has been characterized as the Mediterranean variant .
Fresh fava beans are the major cause of hemolytic anemia, although dried beans and sometimes the pollen of the plant have also been incriminated .
Our data (Table I) are quite different from those described by Kattamis et al.  who observed a higher incidence from dried fava beans. This discrepancy may be related to a greater use of the dried beans in Greece or alternatively to different concentrations of the toxic agents present in the beans. Unlike Schilirò [ 10] and in agreement with Kattamis et al. , we never observed cases of favism from pollen inhalation.
As previously shown , the highest peaks of fava-bean-induced hemolytic anemia were observed during the spring season. Hemolysis in breast-fed babies whose mothers had eaten fava beans has been described by several authors [10-12], but this never occurred in our survey.
Despite the fact that after the first episode of hemolytic anemia several subjects continued to eat fava beans with persisten susceptibility, only 12 experienced irregular episodes of hemolysis. In agreemen with Sartori  and Stamatoyannopoulos et al.  this implies that fava beans can not always trigger off hemolytic anemia and that other additional factors may be involved in the pathogenesis of favism. Stamatoyannopoulos et al.  found familial predisposition to favism and suggested that the familial aggregation of the disease might be due to an additional gene involved in the metabolism or absorptior of the agents present in the beans. Further studies  suggest that favism may be multifactorial syndrome in which both intra- and extraerythrocytic factors concur. With regard to the latter, Cassimos et al.  found that favic subjects had impaired liver capacity in the glucuronization of salicylamide. This aspect was confirmed by Cutillo et al.  who showed a family incidence of the defect. Besides impaired capacity to conjugate salicylamide, reduced urinary excretion of D -glutaric acid has been reported in favic subjects .
With regard to the intraerythrocytic factors associated with favism three major aspects have been considered in the past: the polymorphism of erythrocyte acid phosphatase, the b -thalassemia trait and the blood groups. Concerning the first aspect Bottini et al.  have shown that subjects carrying the P a and P c acid phosphatase phenotypes are probably more prone to develop favism than those with the P b phenotype.
On the contrary, b -thalassemia trait associated with G-6-PD deficiency appears to lower the risk of developing favism. Siniscalco et al.  have shown that only 5 % of the subjects carrying the two genetic defects undergo hemolysis after ingestion of fava beans. Preliminary data [ Meloni et al., unpubl. observation] do not confirm these results since the association of favism and b -thalassemia in the last few years was of the order of 10% which is very close to the incidence of (-thalassemia observed in the population of Sassari. Regarding the third aspect, Hedayat et al.  have reported a higher incidence of favism in subjects carrying the O group. Our findings (Table III) do not confirm their results since the distribution of blood groups in favic subjects are superposable on that of the normal population of Sassari .
Regarding the clinical aspects of favism, anemia and hemoglobinuria associated with general malaise weakness, cephalea and vomiting are the most evident signs; icterus was not always present and when present was not severe. It occurred in relation to the rapidity of hemolysls, the amount of erythrocytes destroyed and the ability of the liver to clear the bilirubin produced by the hemolysis which, as mentioned above, could be genetically determined. The subjects with higher bilirubin values had higher SGPT levels, whereas their SGOT values were no different from those of normal subjects (Table III). These findings contrast with those of Kattamis et al. . They found higher SGOT and SGPT levels during the crisis, attributing them to red cell lysis. Together with increased normal SGOT levels we previously found increased serum levels of y-GT and GLDH . It is our opinion that the slight increase in these enzymatic activities may be related to transient cholestasis .
Table III. Distribution of ABO groups
| ||Number of cases ||O ||A ||B ||AB |
|n ||% ||n ||% ||n ||% ||n ||% |
|Children with favism ||895 ||508 ||56.8 ||288 ||32.2 ||77 ||8.6 ||228 ||2.4 |
|Normal population ||10,786 ||6,042 ||55.9 ||3,387 ||31.3 ||1,111 ||10.3 ||64 ||2.3 |
In conclusion, our study demonstrates that hemolytic anemia occurs frequently in North Sardinia. When one also takes into account the higher incidence of neonatal hyperbilirubinemia among G-6-PD-deficient subjects , it is evident that G6-PD deficiency still represents a serious public health problem. For this reason, a screening program for newborn infants was established some years ago and it is encouraging that such screening has led to a considerable reduction in the number of cases of hemolytic anemia over the last 3 years, only 13 cases having occurred in 1981.
We thank Dr. C.F Gaetani for critical revision of the manuscript and Silvana Sonnu for her excellent editorial assistance.
This work received financial aid by the Consiglio Nazionale delle Ricerche, Progetto Finalizzato Medicina Preventiva, Sottoprogetto MEE.
1. Salvidio, E.; Pannacciulli, I; Tizianello. A.; Gaetani, G.F.; Paravidino, G.: Glucose-6-phosphate dehydrogenase defciency in Italy. Acta haemat. 41: 331-340(1969).
2. Brewer, G.J.: Tarlov, A.R.; Alving, A.S.: The methamoglobin reduction test. A new simple in vitro test for identifying primaquine sensitivity. Bull. Wld Hlth Org. 22: 633-640 (1960).
3. Huisman, T.H.J.: Schroeder, W.A.; Broedie, A.N.; Mayson, S.M.; Jakway, J.: Microchromatography of hemoglobins. III. A simplified procedure for the determination of hemoglobin A 2 . J. Lab. clin. Med. 86: 700-702 (1975).
4. Battistuzzi, G.; Esan, G.J.F.; Fasuan, F.A.; Modiano, G.; Luzzatto, L.: Comparison of GdA and GdB activities in Nigerians. A study of the variation of the G-6-PD activity. Am. J. hum. Genet. 29: 31-36 ( 1977).
5. Sansone, G.; Rasore-Quartino, A.; Veneziano, G.: Dimostrazione su strisci di sangue di una doppia popolazione eritrocitaria nelle donne eterozigoti per la defcienza in glucosio-6-P deidrogenase. Pathologica 55: 371 -375 (1963).
6. Sansone, G.: Piga, A.M.; Segni, G.: Il favismo (Minerva Medica, Torino 1958).
7. Sartori, E.: Elementi per una teoria genetica del favismo.Acta paediat.lat. 10: 506-517(1957).
8. Beutler, E.: Hemolytic anemia in disorders of red cell metabolism; in Wintrobe, Topics in hematology, pp.23- 167 (Plenum Press, New York 1978).
9. Kirkman, H.N.; Schettini, F.; Pickard, B.A.: Mediterranean variant of glucose-6-phosphate dehydrogenase. J. Lab. clin. Med. 63: 726-735 (1965).
10. Schiliro, G.; Russo, A.; Curreri, R.; Marino, S.; Sciotto, A.; Russo, G.: Glucose-6-phosphate deydrogenase defciency in Sicily. Incidence, biochemical characteristics and clinical implications. Clin. Genet. 15: 183-188 (1979).
11. Kattamis, C.A.; Kyriazokou, M.: Chaidas, S.: Favism. Clinical and biochemical data. J. Med. Genet. 6: 34-41 (1969).
12. Kattamis, C.: Favism in breast-fed infants. Archs Dis.Childh. 46: 741 (1971).
13. Stamatoyannopoulos, G.; Fraser, G.R.; Motulski, A.G.; Fessas, P.; Akrivakis, A.; Papayannopoulos, T.: On the familial predisposition to favism. Am. J. hum. Genet. 18: 253-263 (1966).
14. Cassimos, C.; Malaka-Zafiriu, K.; Tsiures, J.; Danielides, B.: Variations in salicylamide glucuronide formation in normal and in G-6-PD deficient children. J. Pediat. 84: 110- 111 (1974).
15. Cutillo, S.; Costa, S.; Vintuleddu, M.C.; Meloni, T.: Salicylamide-glucuronide formation in children with favism and in their parents. Acta haemat.44: 296-299(1976).
16. Cassimos, C.; Malaka-Zafiriu, K.; Tsiuris, J.: D -Glucaric acid excretion in normal and G-6-PD deficient chiIdren. J. Pediat. 84: 871 ( 1974).
17. Bottini, E.; Lucarelli, P.; Agostino, R.; Palmarino, R.; Businco, L.; Antognoni, G.: Favism. Association with erythrocyte acid phosphatase pheno-type. Science 171: 40-411 ( 1971).
18. Siniscalco, M.; Bernini, L.; Latte, B.; Motulski, A.G.: Favism and thalassaemia in Sardinia and their relationship to malaria. Nature, Lond. 190: 1179-1180(1961).
19. Hedayat, S.; Farhud, D.D.; Montazami, K.; Ghadirian, P.: The pattern of bean consumption, laboratory findings in patients with favism, G-6-PD deficient, and a control group. J. trop. Pediat. 27: 110-113 (1981).
20. Stangoni, A.: Marras, G.: Marogna, G.: Frequenza dei gruppi sanguigni ABO e D nella polazione della provincia di Sassari. La T.D.S. 14: 244-247 (1969).
21. Kattamis, C.; Karambula, K.: Ioannidou, V.; Hatzikou, V.: Jaundice and bilirubin levels in Greek children with ravism. Arch Dis. Childh. 51: 233-235 (1976).
22. Meloni, T.; Pilo, C.: Gallisai, D.; Dore, A.: Serum glutamic oxalacetic transaminase, glutamic pyruvic transaminase, gamma-glutamyl transpeptidase and glutamic dehydrogenase levels in favism.Actahaemat.62: 71-73 (1979).
23. Meloni. T.; Forteleoni, G.; Dore, A.; Cutillo, S.: Neonatal hyperbilirubinaemia in heterozygous glucose-6-phosphate dehydrogenase subjects. Br. J. Haemat. (in press).
Received: October 12, 1982
Accepted: February 5, 1983
Clinica Pediatrica dell'Universita di Sassari,
Viale San Pietro,
1-07100 Sassari (Italy)