RESEARCH LETTERS researchers 1 in a subgroup of HIV-1-infected people might be due to other causes, such as decreased synthesis of glutathione. The in-vivo observation of Herzenberg et al 5 on the successful use of N-acetylcysteine in replenishing intracellular GSH in HIV-1-infected people, seems to support this view. We suggest that more than one mechanism may be involved in the metabolism of glutathione in HIV-1-infected cells, with oxidative stress playing a small part. This work was supported by funds from Ricerca Finalizzata Ministero Sanità 1996, RF’96.286/ICS070.1; MURST 1996–1997; and P F Biotecnologie (CNR Target Project on Biotechnology). 1 Staal FJ, Ela SW, Roederer M, Anderson MT, Herzenberg LA, Herzenberg LA. Glutathione deficiency and human immunodeficiency virus infection. Lancet 1992; 339: 909–12. 2 Pace GW, Leaf CD. The role of oxidative stress in HIV disease. Free Radic Biol Med 1995; 19: 523–28. 3 Ferraris AM, Giuntini P, Galiano S, Gaetani GF. 2-deoxy-glucose-6- phosphate utilization in the study of glucose-6-phosphate dehydrogenase mosaicism. Am J Hum Genet 1981; 33: 307–13. 4 Meister A. Biochemistry of glutathione. In: Greenberg DM, ed. Metabolism of sulfur compounds. New York: Academic Press, 1975: 101–88. 5 Herzenberg LA, De Rosa SC, Dubs JG, et al. Glutathione deficiency is associated with impaired survival in HIV disease. Proc Natl Acad Sci USA 1997; 94: 1967–72. Dipartimento di Oncologia Clinica e Sperimentale, Università di Genova; and Istituto Nazionale per la Ricerca sul Cancro, 16132 Genova, Italy (G F Gaetani); Istituto di Malattie Infettive, Università di Sassari fragments generated by PCR. PCR fragments were digested with various endonucleases which cut in or close to putative regulatory elements. The restriction pattern obtained with most of the restriction enzymes was in accordance with the published DNA sequence. 5 A Bsa I-resistant DNA fragment ranging from bp 2727 to 3299 was observed. DNA sequence analysis showed that a point mutation at position -969 (G→C) occurred. Analysis of all PCR fragments by Bsa I restriction showed that there was an unequal distribution of this mutation in the groups (table). There were 30 children with mild malaria heterozygous for this point mutation, and 17 with severe malaria (p=0·040; RR 0·67; 95% CI 0·46–0·96). We found two patients homozygous for the mutation. The distribution of this mutation showed no ethnic bias. In 100 Germans only the wild type occurred. We also determined the frequency of sickle-cell trait, and found a significant difference, in the same range as the distribution of the NOS2 promoter mutation (21 mild cases vs ten severe cases; p=0·045; RR 0·61; 95% CI 0·38–0·96). We compared times to first reinfection. During 9 months of follow-up, 92 patients became reinfected, 24 were lost to follow-up, three severe cases died, and 81 patients were not yet reinfected. Individuals heterozygous for the NOS2- promoter polymorphism had a significantly lower risk of reinfection than the other patients (p=0·022). Differences in times to first reinfection between severe and mild cases and between individuals with and without the sickle-cell trait were not statistically significant. Different genetic mutations have developed which are related to mechanisms that combat malaria. Our data suggest that a new NOS2 promoter mutation is restricted to an area that is highly endemic for malaria. It is likely that this mutation mediates antimalarial resistance through a more effective antiparasitic defence in children who still largely depend on their innate immune response. We thank Ruprecht Schmidt-Ott, Bernhard Greve, Peter Matousek, Klaus Herbich, and Daniela Schmid for patient recruitment and Marcel Nkeyi, Anselme Ndzengue, and Helga Ernst for technical help, and the fortüne programme of the Medical Faculty of the University of Tübingen for financial support. 1 Hill AV. Malaria resistance genes: a natural selection. Trans R Soc Trop Med Hyg 1992; 86: 225–26. 2 Allison AC. Protection afforded by sickle cell trait against subtertian malarial infection. BMJ 1954; 1: 290–94. 3 Kremsner PG, Winkler S, Wildling E, et al. High plasma levels of nitrogen oxides are associated with severe disease and correlate with rapid parasitological and clinical cure in Plasmodium falciparum malaria. Trans R Soc Trop Med Hyg 1996; 90: 44–47. THE LANCET • Vol 351 • January 24, 1998 265 Polymorphism in promoter region of inducible nitric oxide synthase gene and protection against malaria Jürgen F J Kun, Benjamin Mordmüller, Bertrand Lell, Leopold G Lehman, Doris Luckner, Peter G Kremsner In areas hyperendemic for Plasmodium falciparum malaria such as Gabon, children who develop severe malaria and require admission to hospital usually have severe anaemia and hyperparasitaemia. Some children develop severe malaria whereas others are seemingly resistant to it. This prompted the search for factors of natural resistance against severe malaria, 1 the most important being sickle-cell trait. 2 A protective role of nitric oxide (NO) in P falciparum malaria was deduced from the association between plasma concentration of NO and the course of disease. 3 We investigated whether mutations in the control region of the NO-synthase 2 (NOS2) gene could be found in patients with different disease severity. 200 patients from a matched pair, case-control study to compare severe and mild malaria in the Albert Schweitzer Hospital in Lambaréné, Gabon, were investigated as described elsewhere, 4 and were followed up as a longitudinal, prospective study to compare reinfection rates with host genetic polymorphisms. Follow-up examinations were done every 2 weeks. Time to reinfection was defined from admission until the first positive thick blood smear associated with malaria symptoms. Wilcoxon-signed-rank and McNemar tests were applied for paired comparisons. Reinfection data were analysed by Kaplan-Meier estimates and differences between groups were calculated with the Breslow-Gehan-Wilcoxon test. The two patient groups consisted of 61 girls and 39 boys, aged 44 (SE 2) months (table). We looked for mutations in the NOS2 promoter close to putative binding sites for transcription factors by restriction analysis of DNA Severe malaria Mild malaria p (n=100) (n=100) Hyperparasitaemia* (n) 83 0 (>250 000 parasites/μL) Severe anaemia* (n) Haemoglobin <50 g/L 39 0 PCV <25% 73 0 Cerebral malaria (n) 9 0 Hypoglycaemia (n) 8 0 Schizontaemia 24 0 Rectal temperature (ºC) 39·8 (0·1) 39·1 (0·1) p<0·001§ Pulse (beats/min) 130 (2) 120 (2) p<0·001§ Respiratory rate (per min) 41 (2) 35 (1) p=0·005§ Systolic blood pressure (mm Hg) 95 (1) 103 (1) p<0·001§ Vomiting 53 33 p=0·004 Parasitaemia/μL‡ 307 000 (11 500) 10 650 (8500) p<0·001§ PCV (%) 21·6 (0·6) 32·9 (0·4) p<0·001§ Lactate (mmol/L) 2·9 (0·2) 2·4 (0·1) p=0·023§ Sickle cell trait (n) 10 21 p=0·045 NOS2 B/b† (n) 17 30 p=0·040 Data are mean (SE) unless otherwise stated. *Inclusion criteria for severe cases. †Heterozygous individuals for the NOS2 promoter polymorphism. ‡Median, absolute deviation. §Wilcoxon signed rank test. McNemar test. PCV=packed cell volume. Clinical, parasitological, and laboratory findings of patients