Am. J. Trop. Med. Hyg., 65(4), 2001, pp. 335–337 Copyright 䉷 2001 by The American Society of Tropical Medicine and Hygiene
ASSOCIATION OF HELMINTH INFECTION WITH DECREASED RETICULOCYTE COUNTS AND HEMOGLOBIN CONCENTRATION IN THAI FALCIPARUM MALARIA ´ DE ´ RICK GAY, WEERAPONG PHUMRATANAPRAPIN, MATHIEU NACHER, PRATAP SINGHASIVANON, FRE UDOMSAK SILACHAMROON, AND SORNCHAI LOOAREESUWAN Unite´ INSERM 511: Immunobiologie Cellulaire et Mole´culaire des Infections Parasitaires, Faculte´ de me´decine Pitie´-Salpe´trie`re, Paris, France; Faculty of Tropical Medicine, Mahidol University, Hospital for Tropical Diseases, Bangkok, Thailand
Abstract. Following a study showing an association between Ascaris and protection from cerebral malaria, we conducted a cross-sectional study comparing admission hemoglobin concentrations in relation to exposure to helminth infection in 2 separate groups of patients: 111 cerebral malaria cases and 180 mild Plasmodium falciparum malaria cases. Hookworm infections were excluded. Mean hemoglobin concentrations were significantly lower in helminthinfected patients compared to those without helminths, both in the cerebral malaria group (10.1 ⫾ 3 [n ⫽ 47] versus 11.2 ⫾ 2.4 g/dl [n ⫽ 64], P ⫽ 0.04) and the mild malaria group (11 ⫾ 2.5 [n ⫽ 89] vs 12.2 ⫾ 2.7 g/dl [n ⫽ 91], P ⫽ 0.004). Median reticulocyte counts, only available in the cerebral malaria group, were lower in helminth-infected patients compared to those without helminths (15,340/23,760 per l, P ⫽ 0.03). Adjustments for confounders such as body mass index did not alter these associations. These data are consistent with a mechanism causing anemia linked to differences in the immune response of helminth-infected patients during malaria. cioeconomic factors that can potentially influence hemoglobin concentrations, we analyzed the relationship between helminth infection and hemoglobin in the Mon ethnic group alone. The Mon are political and economic refugees residing in the mountains of the western Thai border. Analyzing the relation between helminths and hemoglobin in this homogeneous group was assumed to avoid socioeconomic confounding. All patients had a stool examination by direct examination of a wet preparation of 1.5 mg of stool. Since hookworm is known to be associated with iron deficiency we excluded patients with hookworm infection (135 patients). Since heavy Trichuris trichiura infections have been described as associated with blood loss,6 9 patients with infections of more than 5,000 eggs per gram of stool were excluded. For the subgroup where erythrocyte abnormalities were recorded, a logistic regression model was generated using the presence or absence of anisocytosis as the dependent variable.
INTRODUCTION
Malarial anemia is an important cause of morbidity and mortality.1 The pathophysiology of malarial anemia is still unclear but likely involves destruction of parasitized erythrocytes at schizont rupture, suppression of erythropoiesis by nitric oxide2 and cytokines such as TNF-␣ and IFN-␥,3 and destruction of non-parasitized red blood cells because of reduced red cell deformability.4 We have recently proposed to explain an association between Ascaris infection and protection from cerebral malaria by hypothesizing that endothelial and leukocyte CD23 ligation by IgE-anti IgE immune complexes induce nitric oxide, preventing cytoadherence of parasitized red blood cells.5 Although severe malarial anemia is rare in Thailand, our objective was to see if the magnitude of malarial anemia was influenced by helminth infection. PATIENTS, MATERIALS, AND METHODS
We performed a retrospective study of 2 separate groups of malaria patients hospitalized between 1991 and 1997 at the Hospital for Tropical Diseases, Bangkok. First, we compared hemoglobin concentrations and reticulocyte counts on admission between helminth-infected and non-infected patients in a group of 111 cerebral malaria cases (reticulocyte counts on admission were only available for 61 cerebral malaria cases because they were part of routine clinical testing from 1991 to 1993). Second, we compared hemoglobin concentrations in relation to the presence or absence of helminths in a group of 180 mild P. falciparum malaria cases. The body mass index (BMI: weight/height2) and details of peripheral blood smears were available for 93 mild P. falciparum malaria patients. We did not compare mild malaria and cerebral malaria assuming they were different diseases. Statistical methods. Quantitative values were compared using the ranksum test when distributions were non-Gaussian and the unpaired t-test for unequal variances was used for normally distributed variables. Adjustments for potential confounding variables were performed using multiple linear regression (STATA 6.0). Since helminths are related to so-
RESULTS
There was no significant difference in the median duration of symptoms between helminth-infected and non-infected patients (respectively: median 4/4 days, quartiles 3/3 days and 6/7days). Types of helminths. There were 47 Ascaris lumbricoides infections, 61 moderate Trichuris trichiura infections, and 23 Strongyloides stercoralis infections. Forty-seven patients had single helminth infections, 30 had dual helminth infections, and 10 had 3 triple helminth infections. Hemoglobin and helminths. In both study groups the mean hemoglobin concentration was significantly lower in helminth-infected patients when compared to patients without helminths (Table 1). Adjustments for age, sex, ethnicity, mean corpuscular volume, P. falciparum parasitemia, and symptom duration did not alter the result (Table 1). There was a negative linear trend (P ⫽ 0.01) between the proportion of helminth-infected patients and the level of hemoglobin (Table 2). Similarly, there was a linear trend (P ⫽ 0.03) between the number of
335
336
NACHER AND OTHERS
TABLE 1 Differences in mean hemoglobin concentration between helminthinfected and non-infected patients Mean hemoglobin concentration g/dL (⫾ SD) [number]
Mild malaria Cerebral malaria
Helminth infection*
No helminth infection
11 ⫾ (2.5) [89] 10.1 ⫾ (3) [47]
12.2 ⫾ (2.7) [91] 11.2 ⫾ (2.4) [64]
P value† {Adjusted}‡
P {P P {P
⫽ ⫽ ⫽ ⫽
0.004 0.005} 0.04 0.02}
* Excluding hookworm infections. † Student unpaired t-test. ‡ Adjustments for age, sex, ethnicity, mean corpuscular volume, parasitemia, and symptom duration using multiple linear regression.
infecting helminths and patients with hemoglobin concentrations below 8 g/dL (P ⫽ 0.028) (Table 3). There was a negative correlation between the number of Ascaris eggs and hemoglobin (Spearman’s rho ⫽ ⫺0.15, P ⫽ 0.045), between the number of Trichuris eggs and hemoglobin (Spearman’s rho ⫽ ⫺0.13, P ⫽ 0.07), and between the number of Strongyloides larvae and hemoglobin (Spearman’s rho ⫽ ⫺0.14, P ⫽ 0.06). In the homogeneous Mon ethnic group, helminth-infected patients had lower mean hemoglobin concentrations (n ⫽ 61; 10.9 ⫾ 2.5 g/dL) than those without helminths (n ⫽ 28; 12.4 ⫾ 2.7 g/dL) (P ⫽ 0.009). This remained unchanged after adjusting for age, sex, fever duration, and mean corpuscular volume (P ⫽ 0.015). Other variables linked to anemia. In the model used for adjustments in cerebral malaria (adjusted r-squared 0.45), there was a negative correlation between symptom duration and hemoglobin concentration (coefficient ⫺3.17, P ⬍ 0.001). In the model used for adjustments in mild malaria (adjusted r-squared 0.24), there was a negative correlation between symptom duration and hemoglobin concentration (coefficient ⫺2.21, P ⬍ 0.001), and a positive correlation between the mean corpuscular volume and hemoglobin concentration (coefficient 0.48, P ⫽ 0.02). Adjustments for nutritional status. The association between helminths and low hemoglobin concentration remained significant (P ⫽ 0.006) when adding the BMI and presence/absence of hypochromia as adjustment variables in the subgroup of mild malaria cases where anthropometric measurements and erythrocyte abnormalities were recorded (n ⫽ 93). Within this model (adjusted r-squared ⫽ 0.35), the symptom duration (coefficient ⫺2.72, P ⬍ 0.001) was also negatively correlated with hemoglobin concentration whereas BMI was positively correlated with hemoglobin concentration (coefficient 20610, P ⫽ 0.03). The presence of hypochromia was associated with low hemoglobin concentration (P ⫽ 0.01). Anisocytosis and helminths. Anisocytosis was significantly more frequent in helminth-infected patients (crude odds ratio 2.9 [0.9–8.6]; P ⫽ 0.06). Using anisocytosis as a dependent variable in an unconditional logistic regression model, after adjusting for BMI and evolution duration, helminths were more frequently associated with anisocytosis (odds ratio 4.6 [1.3–16.3]; P ⫽ 0.02). Reticulocytes and helminths. In the cerebral malaria group the median reticulocyte count was lower in helminth-
TABLE 2 Odds for helminth infections for different hemoglobin concentrations in patients with mild Plasmodium falciparum malaria Hemoglobin concentration
Mild malaria with helminth infection
Mild malaria with no helminth infection
6–8 g/dL
11
7
8.1–10 g/dL
21
13
10.1–12 g/dL
47
47
12.1–18 g/dL
10
24
Odds* (CI)
1.4 (0.6–4) 1.6 (0.8–3.2) 1 (0.7–1.5) 0.4 (0.2–0.9)
* Score for linear trend, chi square 6.56 (1 df), P ⫽ 0.01.
infected patients (n ⫽ 24) than in patients without helminths (n ⫽ 37; 15,340 versus 23,760 per L; 25% quartile 9,140— 11,340 per L; 75% quartile 27,180–45,000 per l; P ⫽ 0.03). Adjustments for normalized reticulocyte counts stratified by age, sex, symptom duration, hemoglobin concentration, bilirubin level, parasitemia, mean corpuscular volume, and presence of renal failure did not alter the association (P ⫽ 0.003). DISCUSSION
We have previously observed that helminth infection seems to confer a degree of protection against severe malaria. The present study suggests that helminth infection is associated with significantly lower hemoglobin concentrations and lower reticulocyte counts in patients with both mild and severe malaria. The relation between hemoglobin concentration and the development of cerebral malaria remains unclear. The role of heavy Trichuris6 infections in iron deficiency and anemia has been reported. However, recent studies, one of them in Thailand,7 have not found an association with either Trichuris or Ascaris and anemia.8,9 Therefore, it seems unlikely that the observed difference in hemoglobin concentrations between helminth-infected and non-infected malaria patients was present before malaria occurred. Another possible explanation for these findings could be that the observed hematological difference was linked to preexisting anemia linked to undetected hookworm infections or nutritional abnormalities in helminth-infected patients. The in-
TABLE 3 Odds of having hemoglobin concentration ⬍ 8 g/dL in relation to the number of different helminths infecting the patient Mild malaria with hemoglobin ⬍ 8 g/dL
Mild malaria with hemoglobin ⬎ 8 g/dL
No helminths
5
87
1 helminth species
3
44
2 helminth species
5
25
3 helminth species
2
8
Hemoglobin concentration
* Score for linear trend, chi square 4.81 (1 df), P ⫽ 0.03.
Odds* (CI)
0.06 (0.02–0.14) 0.07 (0.02–0.2) 0.2 (0.07–0.5) 0.25 (0.05–1.2)
337
HELMINTHS AND MALARIAL ANEMIA
sufficient sensitivity of stool examinations could have missed the epidemiological link between hookworm and other helminths. However, adjusting for the body mass index and hypochromia did not alter these findings. Therefore, the differential presence of anemia in these patient groups does not seem to be explained by nutritional factors. It is possible that adjustments for mean corpuscular volume and hypochromia were not sufficient to unmask the confounding effect of iron deficiency, or that the statistical power of this study was limited in being able to detect a weak effect. Possibly, the hematologic stress resulting from malaria unmasked subclinical deficiencies in helminth-infected patients. In accordance with our previous hypothesis5 and subsequent findings showing significantly higher concentrations of reactive nitrogen intermediates (RNI) in helminth-infected malaria patients (Nacher M and others, unpublished data), a possible explanation for the findings of the present study could be that helminth infections lead to decreased erythropoiesis through NO release. Since nitric oxide can reduce erythrocyte deformability,10,11 it could lead to increased red blood cell destruction.4 This hypothesis is supported by recent findings showing that RNI were negatively correlated with hemoglobin concentrations12 and were highest in young infants and after 5 years. Thus RNI could possibly contribute to the epidemiological patterns of severe anemia and cerebral malaria.13 The role of the spleen in anemia in the patient populations studied here remain unclear. Previously, we have observed indirect signs of decreased clearance in helminth-infected malaria patients.5 One hypothesis could be that there is an adaptive, possibly NO-mediated reduction in splenic clearance avoiding selection of cytoadhesive strains.14 This would lead to a delay in crisis-related destruction of parasitized red blood cells and result in higher parasitemia and more profound anemia. In conclusion, further studies are needed to elucidate the underlying mechanisms of the observations presented here. During malaria, preexisting helminth infections are associated with decreased hemoglobin concentrations and reticulocytosis and may aggravate the impact of malarial anemia. The fact that in patients without malaria the same helminth infections are exceptionally correlated with anemia suggests that the effect may be linked to the association of P. falciparum and helminths. Helminth-infected patients may be somewhat protected from cerebral malaria5 but more prone to severe anemia. Thus the prevalence of helminth infections and the intensity of malaria transmission could both provide partial explanations of the mutual exclusivity of severe malaria due to profound anemia versus cerebral complications. Financial support: This work was supported by the French Embassy’s Regional Office for Scientific and Technical Cooperation, Bangkok. Mathieu Nacher is supported by the Lavoisier grant, French Ministry of Foreign Affairs. Authors’ addresses: Mathieu Nacher and Frederick Gay, Unite´ INSERM 511; Immunobiologie Cellulaire et Mole´culaire des Infections Parasitaires, Faculte´ de me´decine Pitie´-Salpe´trie`re, 91 boulevard de l’Hoˆpital, 75634 Paris CEDEX 13, France. Tel: 331 40 77
97 36; Fax: 331 45 82 28 59 email m㛮
[email protected],
[email protected]. Pratap Singhasivanon, Weerapong Phumratanaprapin, Udomsak Silachamroon, and Sornchai Looareesuwan, Faculty of Tropical Medicine, Mahidol University, Hospital for Tropical Diseases, 420/6 Rajvithi Road, 10400 Bangkok, Thailand. Tel (662) 247 16 88, Fax (662) 245 72 88, Email: m㛮
[email protected],
[email protected]. Reprint requests: Mathieu Nacher, Unite´ INSERM 511, Immunobiologie Cellulaire et Mole´culaire des Infections Parasitaires, Faculte´ de me´decine Pitie´-Salpe´trie`re, 91 boulevard de l’Hoˆpital, 75634 Paris CEDEX 13, France. REFERENCES
1. WHO, Division of Control of Tropical Diseases, 1990. Severe and complicated malaria. Trans R Soc Trop Med Hyg 84 Suppl 2: 1S–65S. 2. Maciejewski JP, Selleri C, Sato T, Cho HJ, Keefer LK, Nathan CF, Young NS, 1995. Nitric oxide suppression of human hematopoiesis in vitro-contribution to inhibitory action of interferon-gamma and tumor necrosis factor alfa. J Clin Invest 96: 1085–1092. 3. Miller LH, Good MF, Milon G, 1994. Malaria pathogenesis. Science 264: 1878–1883. 4. Dondorp AM, Angus BJ, Chotivanitch K, Silamut K, Ruangveerayuth R, Hardeman MR, Kager PA, Vreeken J, White NJ, 1999. Red blood cell deformability as a predictor of anemia in severe falciparum malaria. Am J Trop Med Hyg 60: 733–737. 5. Nacher M, Gay F, Singhasivanon P, Krudsood S, Treeprasertsuk S, Mazier D, Vouldoukis I, Looareesuwan S, 2000. Ascaris lumbricoides infection is associated with protection from cerebral malaria. Parasite Immunol 22: 107–115. 6. Layriss M, Aparcedo L, Martinez-torres C, Roche M, 1967. Blood loss due to infection with Trichuris trichiura. Am J Trop Med Hyg 16: 613–619. 7. Trieteeraprab S, Nuchprayoon I, 1998. Eosinophilia, anemia and parasitism in a rural region of northwest Thailand. Southeast Asian J Trop Med Public Health 29: 584–590. 8. Beasley NMR, Tomkins AM, Hall A, Kihamia CM, Lorri W, Nduma B, Issae W, Nokes C, Bundy DAP, 1999. The impact of population level deworming on the haemoglobin level of schoolchildren in Tanga, Tanzania. Trop Med Int Health 4: 744–750. 9. Olsen A, Magnussen P, Ouma JH, Andreassen J, Friis H, 1998. The contribution of hookworm and other parasitic infections to haemoglobin and iron status among children and adults in western Kenya. Trans R Soc Trop Med Hyg 92: 643–649. 10. Korbut R, Gryglewski RJ, 1993. Nitric oxide from polymorphonuclear leukocytes modulates red blood cell deformability in vitro. Eur J Pharmacol 234: 17–22. 11. Starzyk D, Korbut R, Gryglewski RJ, 1997. The role of nitric oxide in regulation of deformability of red blood cells in acute phase of endotoxaemia in rats. J Phys Pharmacol 48: 731– 735. 12. Anstey NM, Granger DL, Hassanali MY, Mwaikambo ED, Duffy PE, Weinberg JB, 1999. Nitric oxide, malaria, and anemia: inverse relationship between nitric oxide production and hemoglobin concentration in asymptomatic, malaria-exposed children. Am J Trop Med Hyg 61: 249–252. 13. Anstey NM, Weinberg JB, Wang Z, Mwaikambo ED, Duffy PE, Granger DL, 1999. Effect of age and parasitemia on nitric oxide production/leukocyte nitric oxide synthase type 2 expression in asymptomatic, malaria-exposed children. Am J Trop Med Hyg 61: 253–258. 14. David PH, Hommel M, Miller LH, Udeinya IJ, Oligino LD, 1983. Parasite sequestration in Plasmodium falciparum malaria: spleen and antibody modulation of cytoadherence of infected erythrocytes. Proc Natl Acad Sci USA 80: 5075– 5079.
