High iron levels are associated with increased malaria risk in infants during the first year of life in Benin

The World Health Organization (WHO) estimates that 40% of children in low-income countries are anemic. Therefore, iron supplements are recommended byWHOin areas with high anemia rates. However, some studies have set into question the benefits of iron supplementation in malaria-endemic regions. In Benin, a west African country with high prevalence of anemia and malaria, no iron supplements are given systematically to infants so far despite the WHO recommendations. In this context, we wanted to investigate the effect of iron levels during the first year of life on malarial risk in Benin considering complementary risk factors. We followed 400 women and their offspring between January 2010 and June 2012 in Allada (Benin). Environmental, obstetric, and numerous clinical, maternal, and infant risk factors were considered. In multilevel models, high iron levels were significantly associated with the risk of a positive blood smear (adjusted odds ratio = 2.90, P < 0.001) and Plasmodium falciparum parasitemia (beta estimate = 0.38, P < 0.001). Infants with iron levels in the lowest quartile were less likely to have a positive blood smear (P < 0.001), and the risk increased with higher iron levels. Our results appeal for additional evaluation of the effect of different doses of iron supplements on the infant health status, including malaria incidence. Thus, the health status of infants should be compared between cohorts where iron is given either for prevention or anemia treatment, to better understand the effect of iron supplements on infant health. A prospective cohort of 400 infants was followed from birth to 12 months of age in the context of the Anemia in Pregnancy: Etiology and Consequences (APEC) study. The APEC study is an ancillary survey nested within the Malaria in Pregnancy Preventive Alternative Drugs (MiPPAD) trial in Benin (http://clinicaltrials.gov/ct2/show/{"type":"clinical-trial","attrs":{"text":"NCT00811421","term_id":"NCT00811421"}}NCT00811421). This study was conducted in three clinics in the district of Allada, between January 2010 and May 2012. Allada is a semirural area of 91,778 inhabitants located 50 km north of Cotonou (Benin). Plasmodium falciparum is the species responsible for the majority of infections. Complete details of MiPPAD are presented elsewhere,13 but briefly, MiPPAD was a randomized trial comparing the efficacy and safety of intermittent preventive treatment in pregnancy (IPTp) with sulfadoxine–pyrimethamine (SP; 1,500/75 mg per dose) and mefloquine (MQ; 15 mg/kg per dose). At delivery, placenta was examined to analyze P. falciparum parasite infestation. Clinical data of the infants were collected during systematic visits at 6, 9, and 12 months of age. In case of sickness, infants were evaluated and, if necessary, treated according to Beninese guidelines (artemether–lumefantrine for uncomplicated malaria; parenteral artesunate or quinine for complicated malaria). In the unscheduled visits, clinical and biological examinations were performed following the same protocol as systematic visits. All drugs prescribed to the infants during the follow-up were free of charge. According to the protocol of the APEC substudy, anthropometric measures and an extensive clinical examination were performed during the visits. In addition, 8 mL of venous blood were collected at each visit. Hemoglobin, serum ferritin, C-reactive protein (CRP), vitamin B12, and folate levels were assessed. A container was also given to the women to collect the infant's stools in search of intestinal helminths. The containers were collected the following day by the study nurses within the first 6 hours after emission. Microbiological examinations were performed as follows: Lambaréné technique was used to assess malaria infection.14 It consists of spreading a calibrated 10 μL amount of blood on a slide's rectangular area of 1.8 cm2 (1.8 × 1 cm). The slide was stained with Giemsa and read at a magnification of 1,000× with an oil immersion lens. To assess parasite density (in parasites/μL), a multiplication factor was applied to the average parasitemia/field. A malaria-infected placenta was defined as a placenta with a positive smear at birth. Helminthic infestations were assessed using the Kato-Katz concentration method (VestergaardFrandsen kit®, New Delhi, India). In case of inflammation (CRP > 5 mg/L), serum ferritin was adjusted following the corrections recommended by Thurnham and others in their meta-analysis,15 to avoid the extrinsic effect of inflammation on serum ferritin levels. More precisely, we multiplied serum ferritin by 0.76 in the presence of Plasmodium without inflammation, and we multiplied serum ferritin by 0.53 in case of concurrent Plasmodium infection and inflammation. We used rain quantity as a surrogate for the risk of exposure to anopheline bites. In the semirural area of Allada, malaria has a perennial transmission pattern with two transmission peaks corresponding to the rainy seasons in April–July and October–November. According to literature, rainfall can be a valid surrogate for anopheline risk.16–18 Because of the anopheline timeliness, rainfall quantity was calculated as the mean rain volume of the 7 days prior to the 2 weeks before the consultation.19 Even if the clinics were close to each other, rainfall quantity was independently assessed for each visit and each clinic. Socioeconomic status was assessed using a socioeconomic index created in a two-step process. First, socioeconomic items (home possession of latrines, electricity, a refrigerator, a television, a vehicle with at least two wheels, the mother being married, and the mother working outside the home) were plotted into a multiple correspondence analysis.20,21 Then, two predictors were created to synthesize the information, and as the first captured the large majority of the information, it was withheld as the socioeconomic index. We used this approach because it allows us to create a synthetic objective index of socioeconomic items without any a priori on the weight of each of the elements of the index. Data were double entered and analyzed with ACCESS 2003, and STATA 12.0 Software (Stata Corp, College Station, TX). Then, exploratory and univariate analyses were performed to assess the association of all variables with both infant positive smear and peripheral P. falciparum density at each visit (systematic or unscheduled visit). χ2 and Kruskal–Wallis tests were used in the univariate analyses. For time-dependent variables, univariate analyses were performed using a random intercept model at the infant level. Then, all variables with P values < 0.2 were included in either a logistic or a linear multivariate multilevel model with a random intercept and slope at the infant level including all visits (systematic and unscheduled visits) for each infant, to explore the determinants of the probability of having a positive smear or peripheral P. falciparum parasitemia, respectively. More precisely, a random slope was applied to the infant age, as the effect of the variables might differ between the infants. The statistical significance in the final multivariate models was set to P < 0.05 (two-sided tests). To evaluate the possible diverse effect of different iron levels on malaria risk, we ran the same multilevel model considering the different quartiles of corrected ferritin. The study was conducted in the context of a clinical trial. According to the International Committee of Medical Journal Editors guidelines, our clinical trial was registered as follows: EDCTP-IP.07.31080.002, MiPPAD study “Malaria in Pregnancy Preventive Alternative Drugs,” (http://clinicaltrials.gov/ct2/show/{"type":"clinical-trial","attrs":{"text":"NCT00811421","term_id":"NCT00811421"}}NCT00811421). This study was approved by the Ethics Committee of the Faculty of Medicine of Cotonou. It was explained in the local language to the mothers and their voluntary consent was obtained before enrollment.