The high frequency of α+-thalassemia in malaria-endemic regions may reflect natural selection due to protection from potentially fatal severe malaria. In Africa, bearing 90% of global malaria morbidity and mortality, this has not yet been observed. We tested this hypothesis in an unmatched case-control study among 301 Ghanaian children with severe malaria and 2107 controls (62% parasitemic). In control children, α+-thalassemia affected neither prevalence nor density of Plasmodium falciparum. However, heterozygous α+-thalassemia was observed in 32.6% of controls but in only 26.2% of cases (odds ratio [OR], 0.74; 95% confidence interval [CI], 0.56-0.98). Protection against severe malaria was found to be pronounced comparing severe malaria patients with parasitemic controls (adjusted OR in children < 5 years of age, 0.52; 95% CI, 0.34-0.78) and to wane with age. No protective effect was discernible for homozygous children. Our findings provide evidence for natural selection of α+-thalassemia in Africa due to protection from severe malaria.

Severe Plasmodium falciparum malaria is a major cause of death in sub-Saharan Africa, particularly in children younger than 5 years of age.1  Interindividual variation in susceptibility and manifestation can be attributed partially to innate host factors such as the sickle cell trait.2-4  According to Haldane's “malaria hypothesis,”5  these factors are subject to selection in endemic regions because of the resistance they confer against malaria. Another hemoglobin disorder, α+-thalassemia, reaches a prevalence of more than 80% in parts of Southeast Asia and Melanesia.6-8  Heterozygosity is characterized by the commonly deletional loss of one of the duplicated α-globin genes (–α/αα) and slight hematologic changes. Homozygous individuals (–α/–α) have mild microcytic anemia.9 

In Melanesia, the geographic correlation between the prevalence of α+-thalassemia and malarial endemicity suggests natural selection.6  In one case-control study in Papua New Guinea, the risk of severe malaria was reduced by 60% and 34% in homozygous and heterozygous children, respectively.7  In sub-Saharan Africa, α+-thalassemia affects up to 50% of the population, but protection against uncomplicated or severe malaria could not be demonstrated so far.3,9,10 

We examined whether α+-thalassemia confers protection from severe falciparum malaria in Ghana.

Patients and controls

We conducted an unmatched case-control study in the city of Tamale and its vicinity, northern Ghana, during the rainy season 2002 (August to November). In the study area, malaria is hyperendemic (F.P.M., unpublished observations, December 1, 2003). Ethical clearance was obtained from the Ethics Committee, University for Development Studies, Tamale, and informed consent, from the participants' parents. We recruited 290 children with severe malaria according to WHO definition at Tamale Teaching Hospital.1,11  An additional 11 children with severe malaria were detected during the survey mentioned later in this paragraph and classified as cases. In the 301 index patients, symptoms defining severe malaria were as follows: severe anemia, 57%; prostration, 32%; respiratory distress, 22%; multiple convulsions, 20%; impaired consciousness, 19%; jaundice, 11%; circulatory collapse, 3%; and hemoglobinuria, 3%. Controls were 2107 children from Tamale and the surrounding districts who were recruited following a 2-stage cluster sampling strategy with probability proportional to population size. There were 30 census units included, and within each 70 or more children aged 6 months to 9 years were randomly selected. Due to ethnic conflicts, ethnicity could not be dependably assessed, but the vast majority of cases and controls belonged to the Dagomba ethnic group. Severe malaria patients received artesunate (Plasmotrim; Mepha, Basel, Switzerland) for 5 days (5 mg/kg body weight, double dose on first day) and supportive care. Uncomplicated malaria or parasitemia higher than 5000/μL was treated with sulfadoxine-pyrimethamine.

Microscopy and PCR assays

Venous blood was taken and DNA extracted (QIAmp; Qiagen, Hilden, Germany). Malaria parasites were counted per 200 or more white blood cells on Giemsa-stained thick blood films. Polymerase chain reaction (PCR) assays were applied to ascertain P falciparum and to determine the common West African –α3.7 deletion type of α+-thalassemia.12,13 

Statistical analysis

Geometric mean parasite densities (GMPDs) and 95% confidence intervals (95% CIs) were calculated. Continuous variables were compared by t test, analysis of variance, Mann-Whitney U test, and Kruskal-Wallis test, and proportions were compared by χ2-tests. The odds of severe malaria in children with α+-thalassemia were estimated in logistic regression models adjusting for sex, age (months), and residence (urban, Tamale district; rural, surrounding districts).

In control children, P falciparum occurred in 62% and α+-thalassemia occurred in 37% (Table 1). The α-globin genotype did not influence P falciparum prevalence (αα/αα, 815 [61%] of 1335 children; –α/αα, 430 [63%] of 686 children; –α/–α, 51 [59%] of 86 children), febrile parasitemia (αα/αα, 58 [4.4%] of 1332 children; –α/αα, 38 [5.6%] of 684 children; –α/–α, 5 [5.8%] of 86 children), or parasite density (GMPDs [95% CIs]: 1500/μL [1313-1712]; 1282/μL [1053-1562]; 1409/μL [807-2462]).

Table 1.

Characteristics of control group children and severe malaria patients



Control group

Severe malaria patients

All
Aparasitemic children
Parasitemic children
No.   2107   811   1296   301  
Age, mo, median (range)   50 (6-108)*  48 (6-108)*  60 (6-108)*  24 (6-102)  
Proportion younger than 5 y of age, no. (%)   1093 (51.9)*  479 (59.1)*  614 (47.4)*  261 (86.7)  
Female, no. (%)   1050 (49.8)   400 (49.3)   650 (50.2)   159 (52.8)  
Rural residence, no. (%)   1353 (64.2)*  326 (40.2)   1027 (79.2)*  104 (34.6)  
Geometric mean parasite density, per μL (95% CI)   1419 (1274-1581)*  0   1419 (1274-1581)*  27 164 (20 246-36 447)  
Heterozygous α+-thalassemia, no. (%)   686 (32.6)   256 (31.6)   430 (33.2)   79 (26.2) 
Homozygous α+-thalassemia, no. (%)
 
86 (4.1)
 
35 (4.3)
 
51 (3.9)
 
15 (5.0)
 


Control group

Severe malaria patients

All
Aparasitemic children
Parasitemic children
No.   2107   811   1296   301  
Age, mo, median (range)   50 (6-108)*  48 (6-108)*  60 (6-108)*  24 (6-102)  
Proportion younger than 5 y of age, no. (%)   1093 (51.9)*  479 (59.1)*  614 (47.4)*  261 (86.7)  
Female, no. (%)   1050 (49.8)   400 (49.3)   650 (50.2)   159 (52.8)  
Rural residence, no. (%)   1353 (64.2)*  326 (40.2)   1027 (79.2)*  104 (34.6)  
Geometric mean parasite density, per μL (95% CI)   1419 (1274-1581)*  0   1419 (1274-1581)*  27 164 (20 246-36 447)  
Heterozygous α+-thalassemia, no. (%)   686 (32.6)   256 (31.6)   430 (33.2)   79 (26.2) 
Homozygous α+-thalassemia, no. (%)
 
86 (4.1)
 
35 (4.3)
 
51 (3.9)
 
15 (5.0)
 
*

Difference to severe malaria patients, P < .0001.

Difference between aparasitemic and parasitemic control children, P < .0001.

Heterozygous α+-thalassemia occurs less frequently in severe malaria patients than in all controls (P = .03) and than in parasitemic control children (P = .02).

We compared the frequencies of α+-thalassemia in cases and controls (Table 1). Cases and controls differed with respect to residence, but heterozygous and homozygous α+-thalassemia occurred at similar prevalence in rural (477 [32.7%] of 1457 children and 62 [4.3%] of 145 children, respectively) and urban (288 [30.3%] of 951 children and 39 [4.1%] of 951 children, respectively; P = .4) areas. Children with heterozygous α+-thalassemia had reduced odds of severe malaria when compared with all controls (odds ratio [OR], 0.74; 95% CI, 0.56-0.98; P = .03) or with parasitemic controls (OR, 0.72; 95% CI, 0.54-0.97; P = .03). These associations were stronger in children younger than 5 years, and strongest when parasitemic children were chosen as the control group (Table 2). In these, logistic regression analysis revealed that heterozygous α+-thalassemia reduced the odds of severe malaria by 48% (Table 3). Further age stratification showed a waning protective effect with increasing age (ORs [95% CIs]: 0.5-2 years, 0.48 [0.28-0.81]; 3-4 years, 0.60 [0.31-1.18]; 5-6 years, 1.55 [0.65-3.72]; ≥ 7 years, 2.01 [0.60-6.75]). In the logistic regression model, the interaction between age and heterozygous α-thalassemia, (ie, the change of effect with age) was significant (P = .0005). Irrespective of age, homozygous α+-thalassemia did not reduce the odds of severe malaria.

Table 2.

Distribution of α-globin genotypes in Ghanaian children with severe malaria and in control groups




Genotype frequencies (no., %)
Stratum
No.
Normal α-globin genotype
Heterozygous α+-thalassemia
Homozygous α+-thalassemia
Younger than 5 years of age     
    Severe malaria   261   186 (71.3)   62 (23.7)   13 (5.0)  
    All controls   1093   699 (63.9)   357 (32.7)   37 (3.4)  
    Aparasitemic controls   479   309 (64.5)   153 (31.9)   17 (3.6)  
    Parasitemic controls   614   390 (63.5)   204 (33.2)   20 (3.3)  
5 years of age or older     
    Severe malaria cases   40   21 (52.5)   17 (42.5)   2 (5.0)  
    All controls   1014   636 (62.7)   329 (32.5)   49 (4.8)  
    Aparasitemic controls   332   211 (63.6)   103 (31.0)   18 (5.4)  
    Parasitemic controls
 
682
 
425 (62.3)
 
226 (33.1)
 
31 (4.6)
 



Genotype frequencies (no., %)
Stratum
No.
Normal α-globin genotype
Heterozygous α+-thalassemia
Homozygous α+-thalassemia
Younger than 5 years of age     
    Severe malaria   261   186 (71.3)   62 (23.7)   13 (5.0)  
    All controls   1093   699 (63.9)   357 (32.7)   37 (3.4)  
    Aparasitemic controls   479   309 (64.5)   153 (31.9)   17 (3.6)  
    Parasitemic controls   614   390 (63.5)   204 (33.2)   20 (3.3)  
5 years of age or older     
    Severe malaria cases   40   21 (52.5)   17 (42.5)   2 (5.0)  
    All controls   1014   636 (62.7)   329 (32.5)   49 (4.8)  
    Aparasitemic controls   332   211 (63.6)   103 (31.0)   18 (5.4)  
    Parasitemic controls
 
682
 
425 (62.3)
 
226 (33.1)
 
31 (4.6)
 
Table 3.

Associations with severe malaria in α-thalassemic children younger than 5 years of age




Heterozygous α+-thalassemia

Homozygous α+-thalassemia
Univariate analysis (OR, 95% CI)   
    Severe malaria vs all controls   0.65 (0.47-0.90); P = .008   1.32 (0.63-2.60); P = .40  
    Severe malaria vs aparasitemic controls   0.67 (0.47-0.96); P = .02   1.27 (0.55-2.85); P = .53  
    Severe malaria vs parasitemic controls   0.64 (0.45-0.90); P = .008   1.36 (0.61-2.95); P = .40  
Multivariate analysis (OR, 95% CI)   
    Severe malaria vs all controls   0.71 (0.51-0.99); P = .04   1.47 (0.73-2.96); P = .28  
    Severe malaria vs aparasitemic controls   0.75 (0.52-1.07); P = .11   1.22 (0.57-2.63); P = .60  
    Severe malaria vs parasitemic controls
 
0.52 (0.34-0.78); P = .002
 
1.87 (0.76-4.62); P = .17
 



Heterozygous α+-thalassemia

Homozygous α+-thalassemia
Univariate analysis (OR, 95% CI)   
    Severe malaria vs all controls   0.65 (0.47-0.90); P = .008   1.32 (0.63-2.60); P = .40  
    Severe malaria vs aparasitemic controls   0.67 (0.47-0.96); P = .02   1.27 (0.55-2.85); P = .53  
    Severe malaria vs parasitemic controls   0.64 (0.45-0.90); P = .008   1.36 (0.61-2.95); P = .40  
Multivariate analysis (OR, 95% CI)   
    Severe malaria vs all controls   0.71 (0.51-0.99); P = .04   1.47 (0.73-2.96); P = .28  
    Severe malaria vs aparasitemic controls   0.75 (0.52-1.07); P = .11   1.22 (0.57-2.63); P = .60  
    Severe malaria vs parasitemic controls
 
0.52 (0.34-0.78); P = .002
 
1.87 (0.76-4.62); P = .17
 

Children with a normal α-globin genotype serve as reference groups. Multivariate ORs are derived from logistic regression models with severe malaria as outcome variable and sex, age (months), and residence (urban/rural) as independent variables. In children 5 years of age or older, α-globin genotypes did not differ significantly between cases and controls.

In severe malaria patients, GMPDs were virtually identical in α-globin normal, heterozygous, and homozygous children (GMPD [95% CI]: 27 797/μL [19 532-39 560], 25 763/μL [14 485-45 821], and 26 182/μL [6194-110 665], respectively). Also, heterozygous and homozygous α+-thalassemia occurred at similar frequencies in the 2 predominant syndromes, that is, severe anemia without cerebral involvement (n = 99; –α/αα, 25.3%; –α/–α, 3.0%) and cerebral involvement (prostration, convulsions, or impaired consciousness, n = 186; 27.4%, 6.5%, P = .4). In 285 hospitalized severe malaria patients with available follow-up data, fatality rates did not differ significantly between children with a normal α-globin genotype (21 [10.9%] of 193 children), heterozygous α+-thalassemia (10 [12.8%] of 78 children), and homozygous α+-thalassemia (1 [7.1%] of 14 children).

We show for the first time that heterozygous α+-thalassemia protects from severe malaria in African children. Our findings are different from those in New Guinea where protection against severe malaria was stronger in homozygous than in heterozygous children.7  We could not detect protection in homozygotes, which is possibly due to their small proportion and the statistical limitations associated. Considering the high malaria burden in sub-Saharan Africa, the comparatively lower frequency of α+-thalassemia in Ghana could indicate a balanced polymorphism.14  Yet, there is no evidence for a reduced biologic fitness of homozygous individuals, or for such to be present particularly in Africa. Close-to-fixation frequencies of α+-thalassemia in Asia and Melanesia suggest additional selecting factors (eg, an increased susceptibility to potentially beneficial Plasmodium vivax infections).15  Also, α+-thalassemia in New Guinea conferred protection against common nonmalaria diseases (eg, respiratory infections).7  In West Africa, P vivax is virtually absent and α+-thalassemic resistance to nonmalaria illness has not been reported.

Regarding possible confounding due to other protective traits or ethnicity, no correlation of α-globin genotypes with hemoglobin C, hemoglobin S, or glucose-6-phosphate dehydrogenase deficiency was observed in a subset of children (F.P.M., unpublished observations, December 1, 2003). Gene frequencies in children of rural or urban origin were very similar. Considering that ethnicity may determine residence, and vice versa, this renders ethnic bias unlikely. Moreover, in contrast to HbS, α-globin genotype frequencies in Ghana are remarkably stable in geographic terms, that is, they are almost identical in northern and southern parts of the country.16 

The mechanism of protection in α+-thalassemia remains obscure. Consistent with previous findings,7,9,10  α+-thalassemia did not influence susceptibility to infection per se. Rather, the heterozygous trait ameliorated disease manifestation, that is, it influenced progression from parasitemia to severe malaria. However, fatality was unchanged, indicating an abrogation of protection once a critical stage of disease is reached. P falciparum–infected α-thalassemic red cells show a decreased capacity to form rosettes, a pathogenic marker of severe malaria,17  but in vitro results on inhibited parasite multiplication are conflicting.18-21  Increased antibody binding to parasitized thalassemic red cells suggests raised neoantigen expression, which would result in enhanced immune recognition and increased parasite clearance.21  The protective effect of heterozygous α+-thalassemia was strongest in the youngest children, that is, during a time when malaria-associated mortality is highest. Similarly, protective effects of the sickle cell trait against severe and uncomplicated malaria in Kenyan and Nigerian children, respectively, have been reported to be restricted to young age groups.2,4  These observations agree with the notion that innate resistance acts mainly before specific immunity has developed.14  The restriction of heterozygous protection to young children could thus indicate a predominance of nonimmune mechanisms, or, alternatively, an accelerated acquisition of immunity. Development of immunity with age in α-globin normal children could explain the decline in heterozygous advantage, but the reasons for the indication of its reversal remain unclear. Older children were rare among severe malaria patients and, thus, this observation requires verification. Irrespective of its underlying causes, protection in young heterozygous children improves the chance of survival and, thus, of reaching an age at which the physical capacity to cope with the infection is higher. At any rate, the number of severe malaria cases prevented among young heterozygous children clearly outweigh a yet-to-be-confirmed risk increase in older children.

In Melanesia, increased susceptibility to P vivax in young α+-thalassemic children has been suggested to induce cross-reactive immune protection against subsequent falciparum malaria.15  Preliminary data indicate no likewise excess of Plasmodium ovale or Plasmodium malariae in α+-thalassemic children in Ghana (F.P.M., unpublished observations, December 1, 2003).

Our results demonstrate that heterozygous α+-thalassemia confers protection against severe malaria. This increased biologic fitness may have contributed to its high frequency in sub-Saharan Africa.

Prepublished online as Blood First Edition Paper, June 22, 2004; DOI 10.1182/blood-2003-11-4090.

Supported by Charité grant 2003-676.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 U.S.C. section 1734.

We thank the children and their parents who participated in this study. We are grateful to the members of the Northern Region Malaria Project (NORMAP) for patient recruitment and to Baerbel Jakob and Susanne Roewer for technical assistance.

1
WHO. Severe falciparum malaria.
Trans R Soc Trop Med Hyg.
2000
;
94
(suppl 1)
1
-90.
2
Aidoo M, Terlouw DJ, Kolczak MS, et al. Protective effects of the sickle cell gene against malaria morbidity and mortality.
Lancet.
2002
;
359
:
1311
-1312.
3
Lell B, May J, Schmidt-Ott RJ, et al. The role of red blood cell polymorphisms in resistance and susceptibility to malaria.
Clin Infect Dis.
1999
;
28
:
794
-799.
4
Guggenmoos-Holzmann I, Bienzle U, Luzzatto L. Plasmodium falciparum malaria and human red cells, II: red cell genetic traits and resistance against malaria.
Int J Epidemiol.
1981
;
10
:
16
-22.
5
Haldane JBS. The rate of mutation of human genes.
Hereditas
1949
;
35
(suppl 1):
267
-272.
6
Flint J, Hill AV, Bowden DK, et al. High frequencies of alpha-thalassaemia are the result of natural selection by malaria.
Nature.
1986
;
321
:
744
-750.
7
Allen SJ, O'Donnell A, Alexander ND, et al. α+-Thalassemia protects children against disease caused by other infections as well as malaria.
Proc Natl Acad Sci U S A.
1997
;
94
:
14736
-14741.
8
Modiano G, Morpurgo G, Terrenato L, et al. Protection against malaria morbidity: near-fixation of the α-thalassemia gene in a Nepalese population.
Am J Hum Genet.
1991
;
48
:
390
-397.
9
Mockenhaupt FP, Bienzle U, May J, et al. Plasmodium falciparum infection: influence on hemoglobin levels in alpha-thalassemia and microcytosis.
J Infect Dis.
1999
;
180
:
925
-928.
10
Allen SJ, Rowe P, Allsopp CE, et al. A prospective study of the influence of alpha thalassaemia on morbidity from malaria and immune responses to defined Plasmodium falciparum antigens in Gambian children.
Trans R Soc Trop Med Hyg.
1993
;
87
:
282
-285.
11
Mockenhaupt FP, Ehrhardt S, Burkhardt J, et al. Manifestation and outcome of severe malaria in children in northern Ghana.
Am J Trop Med Hyg.
2004
;
71
:
XXX
-XXX.
12
Djimde A, Doumbo OK, Cortese JF, et al. A molecular marker of chloroquine-resistant falciparum malaria.
N Engl J Med.
2001
;
344
:
257
-262.
13
Liu YT, Old JM, Miles K, Fisher CA, Weatherall DJ, Clegg JB. Rapid detection of alpha-thalassaemia deletions and alpha-globin gene triplication by multiplex polymerase chain reactions.
Br J Haematol.
2000
;
108
:
295
-299.
14
Allison AC. Polymorphism and natural selection in human populations.
Cold Spring Harb Symp Quant Biol.
1964
;
29
:
137
-149.
15
Williams TN, Maitland K, Bennett S, et al. High incidence of malaria in alpha-thalassaemic children.
Nature.
1996
;
383
:
522
-525.
16
Mockenhaupt FP, Ehrhardt S, Otchwemah R, et al. Limited influence of haemoglobin variants on Plasmodium falciparum msp1 and msp2 alleles in symptomatic malaria.
Trans R Soc Trop Med Hyg.
2004
;
98
:
302
-310.
17
Carlson J, Nash GB, Gabutti V, al-Yaman F, Wahlgren M. Natural protection against severe Plasmodium falciparum malaria due to impaired rosette formation.
Blood.
1994
;
84
:
3909
-3914.
18
Luzzi GA, Torii M, Aikawa M, Pasvol G. Unrestricted growth of Plasmodium falciparum in microcytic erythrocytes in iron deficiency and thalassaemia.
Br J Haematol.
1990
;
74
:
519
-524.
19
Senok AC, Li K, Nelson EA, Arumanayagam M, Li CK. Flow cytometric assessment of oxidant stress in age-fractionated thalassaemic trait erythrocytes and its relationship to in vitro growth of Plasmodium falciparum.
Parasitology.
1998
;
116
:
1
-6.
20
Pattanapanyasat K, Yongvanitchit K, Tongtawe P, et al. Impairment of Plasmodium falciparum growth in thalassemic red blood cells: further evidence by using biotin labeling and flow cytometry.
Blood.
1999
;
93
:
3116
-3119.
21
Luzzi GA, Merry AH, Newbold CI, Marsh K, Pasvol G, Weatherall DJ. Surface antigen expression on Plasmodium falciparum–infected erythrocytes is modified in α- and β-thalassemia.
J Exp Med.
1991
;
173
:
785
-791.
Sign in via your Institution