Menu
Menu
Menu
Menu
Background Anaemia is an increasingly recognized health problem in Africa, particularly in infants and pregnant women. Although malaria is known to be the main risk factor of anaemia in both groups, the consequences of maternal factors, particularly malaria in pregnancy (MiP), on infant haemoglobin (Hb) concentrations during the first months of life are still unclear. Methods We followed-up a cohort of 1005 Beninese pregnant women from the beginning of pregnancy until delivery. A subsample composed of the first 400 offspring of these women were selected at birth and followed until the first year of life. Placental histology and blood smear at 1st clinical antenatal visit (ANC), 2nd ANC and delivery were used to assess malaria during pregnancy. Infant Hb concentrations were measured at birth, 6, 9 and 12 months of age. A mixed multi-level model was used to assess the association between MiP and infant Hb variations during the first 12 months of life. Results Placental malaria (difference mean [dm] = – 2.8 g/L, 95% CI [-5.3, -0.3], P = 0.03) and maternal peripheral parasitaemia at delivery (dm = – 4.6 g/L, 95% CI [-7.9, -1.3], P = 0.007) were the main maternal factors significantly associated with infant Hb concentrations during the first year of life. Poor maternal nutritional status and malaria infection during infancy were also significantly associated with a decrease in infant Hb Conclusion Antimalarial control and nutritional interventions before and during pregnancy should be reinforced to reduce specifically the incidence of infant anaemia, particularly in Sub-Saharan countries.
Prospective cohort study, set-up between January 2010 and May 2012 to follow pregnant women from the first antenatal clinical visit (1st ANC) to delivery in the framework of the MiPPAD clinical trial. The first 400 infants to be delivered from these women, were enrolled from January 2010 until June 2011 and followed throughout the first year of life. The study was conducted in two maternity clinics (Attogon and Sékou) in the district of Allada, a semi-rural area located 50 km north of Cotonou and the economic capital of Benin. Allada district is characterized by a subtropical climate and malaria is hyperendemic with an average of 20.5 infected anopheles bites/person/year [9]. Plasmodium falciparum is the predominant species transmitted (97%). The study population was composed of HIV—negative pregnant women and their children residing in the district of Allada. Twin pregnancies, pregnancies complicated by stillbirth or fetal abnormalities have been excluded. IPTp was administered according to the clinical trial protocol and the details are presented elsewhere [8]. Briefly, two doses of IPTp (1,500/75 mg of sulfadoxine-pyrimethamine or 15 mg/kg of mefloquine per dose) were administered at ANC visits. The second dose of IPTp was given at least 1 month apart from the first dose. Clinical malaria episodes were treated with oral quinine or arthemether-lumefantrine in first and subsequent trimesters, respectively, for uncomplicated malaria, and with parenteral quinine in case of severe malaria. Children with clinical malaria were also treated with arthemether-lumefantrine according to the national malaria control program guidelines and in case of severe malaria, they were referred to the district hospital. Each woman received a long-lasting insecticide-treated net that was replaced in case of damage or loss during the follow-up. They were also systematically given 600 mg of mebendazole at 1st ANC to be taken at home (100 mg twice day for 3 days) according to the guidelines of the Beninese Ministry of Health. In addition, supplements of oral ferrous sulfate (200 mg/day) and folic acid (5 mg/day) were given to the pregnant women to be taken at home. Pregnant women with a Hb concentration below 110g/L were treated according to the severity of anaemia (i.e., 200 mg oral ferrous sulfate twice day for mild and moderate anaemia (Hb level between 70 and 110 g/L) and referred to the district tertiary hospital in case of severe anaemia (Hb < 70 g/L)). Mild or moderate infant anaemia was treated with 10 mg/kg/day of iron syrup for two months. Children with severe anaemia were also referred to the district tertiary hospital for blood transfusion. Infant with positive stool test were treated with mebendazole syrup (100 mg twice day for 3 days). All the medications prescribed to the women and the children during their participation in the study were free of charge. After obtaining informed consent, we collected sociodemographic and socioeconomic characteristics of participants at enrolment. At the 1st ANC, women were examined and gestational age, middle upper arm circumference (MUAC), weight and height were recorded. This information, except for height, was also collected at 2nd ANC and delivery. Gestational age was determined from fundal height measurement by bimanual palpation and following McDonald's rules [10]. Weight and height in pregnant women were respectively measured to the nearest 0.1 kg using an electronic scale (Seca corp., Hanover, MD) and to the nearest 0.1 cm by using a bodymeter device (Seca 206 Bodymeter; Seca corp.). These parameters were measured by two nurses, and the mean of two measurements was calculated. At birth, newborn’s sex, weight, length, head circumference and axillary temperature were collected. Weight was measured using an electronic baby scale (SECA type 354) with a precision of 10 g and length was measured to the nearest 1 mm with a locally manufactured wooden measuring scale according to the criteria recommended by WHO. At the 6, 9 and 12 months scheduled visits, the history of fever within the previous 24 hours, malaria treatment or hospitalization since the last visit and use of insecticide-treated nets were recorded (Fig 1). During follow-up, socio-demographic, economic, clinical and biological data were collected in mothers at 1st antenatal clinical visit (ANC), 2nd ANC and delivery. The same data were also recorded in infants at birth, 6, 9 and 12 months of life. Outside of scheduled visit, haemoglobin concentration and blood smear were performed when malaria signs were present. At 1st ANC, 2nd ANC and delivery, 8 mL of mother’s venous blood was collected. The same volume was also collected on cord blood at birth and on infant's venous blood at 6, 9 and 12 months of life. All these samples were used to look for malaria parasitaemia, to determine C-reactive protein (CRP), micronutrient (serum ferritin, folic acid and vitamin B12) and Hb concentration and to genotype Hb. At delivery, samples (biopsy and impression smear) were collected from the placenta for parasitological evaluation. A container was also given to the woman to collect infant’s stools in search of intestinal helminths. On unscheduled visits, Hb dosages and thick blood smears were performed in infants with clinical signs of malaria (history of fever in the last 24 hours or temperature ≥ 37.5°C and pallor). The Hb level was measured with a Hemo-Control photometer (EKF Diagnostics, Magdeburg, Germany) device. Daily calibration of the Hemo-Control device was performed by laboratory technicians. In addition, an external quality control was made by sending one of 10 consecutive samples to the Allada Central Hospital laboratory, where dosages were assessed using a hematology analyzer (Erma Laboratory, Tokyo, Japan). Hb genotypes were determined by alkaline electrophoresis on cellulose acetate (Helena laboratories, Beaumont, TX). Serum ferritin, folic acid, and vitamin B12 concentrations were measured using a microparticle enzyme and fluorescence polarization immunoassay (AxSym Immuno-Assay Analyzer, Abbott Laboratories). CRP concentration was determined by rapid slide test (CRP Latex; Cypress Diagnostics Inc.) to correct the effect of inflammatory syndromes on ferritin concentrations. The Determine (HIV1 and 2 kit; Abbott Laboratories) and Bioline (HIV1 and 2 3.0 kit; Bioline, Taunton, MA) rapid tests were used to detect HIV infections using a serial testing algorithm. The Lambaréné technique was used to analyze peripheral malaria infection in blood smears. It consists of spreading a calibrated 10 μL amount of blood on a slide’s rectangular area of 1.8 cm² (1.8 x 1 cm). The slide was stained with Giemsa and read at a magnification of 1,000 × with an oil immersion lens. A multiplication factor was applied to the average parasitemia/field to determine the number of parasites/μL. The Lambaréné method detection threshold has been estimated to be 5 parasites/μL [11,12]. Placental biopsies (2.5 x 2.5 cm3), collected at delivery for histology assessment, were immediately put in 50 ml of 10% buffered formalin. It was then stored at 4°C in a refrigerator until the placental tissue was processed at the pathology department. The maximum delay before fixation was of 5 days. Placental malaria infection was defined as the presence of parasites with /without pigment or pigment confined to fibrin in the histological examination [13]. Placental histology was examined without knowledge of the peripheral blood smears results. In addition, an external quality control was made on 100% of positive slide and 10% of negative slide in reference laboratory to Barcelona Centre for International Health Research (CRESIB), Hospital Clínic-Universitat de Barcelona. Infestations by helminths were assessed by using the Kato-Katz concentration method (Vestergaard Frandsen, Lausanne, Switzerland). Anaemia. Gestational anaemia was defined as a Hb concentration below 110 g/L [14]. In children, it was defined by Hb level below 140 g/L at birth [15] and below 110 g/L from 6 months [14]. Severe, moderate, and mild anaemia were respectively defined as Hb concentrations less than 80 g/L, between 80 and 99 g/L, and between 100 and 109 g/L. Maternal anaemia was also assessed by number of anaemia episodes during pregnancy and in purpose pregnant women have been classified into 4 groups: (1) women not presenting any episode of anaemia during pregnancy (either at inclusion, at the second antenatal visit or at delivery); (2) women presenting a single episode; (3) women with two episodes and (4) pregnant women constantly anaemic (anaemia at each of the 3 blood assessments). Iron status and iron deficiency anaemia (IDA). Iron deficiency (ID) was defined as a serum ferritin concentration less than 12 μg/L or as serum ferritin concentration of 12–70 μg/L in the context of inflammatory syndrome (CRP concentration ≥ 6 mg/mL) [16]. Anaemia associated to ID determined IDA. Folic acid and vitamin B12 deficiencies. Folic acid and vitamin B12 deficiencies were respectively defined as a serum folic acid concentration less than 6 ng/mL and a vitamin B12 serum concentration less than 150 pg/mL [17]. Helminth infestations. Intestinal helminth infestations were diagnosed by the presence of intestinal helminth eggs in the stool sample. Eggs were counted as number of eggs per g of stool. Estimation of pre-pregnancy body mass index (BMI). All pregnant women included in the study had a gestational age less than 28 weeks. From the end of the first trimester of gestation, it was estimated that pregnant women gained on average 1 kg per month until delivery [18]. We used the gestational age at inclusion to estimate approximately the weight that women were supposed to have gained since the beginning of the pregnancy. This amount was then subtracted from the weight on the day of inclusion to obtain a rough estimate of the weight before pregnancy. BMI was calculated as the weight in kilograms divided by the square of the height in meters (kg/m²) and we used WHO criteria to classify the women in normal BMI, thinness (severe, moderate, mild), overweight and obesity [19]. Socio-economic index of family. This index was obtained from a synthetic score given by the sum of the following binary variables: having gainful activity, electricity, television, refrigerator and at least a bicycle. The total score was graded between 0 and 5. Low, average, high socio-economic indexes were defined as synthetic score ≤ 1, between 2–3, and ≥ 4, respectively. Data were double-entered into Microsoft Access 2003 database and analyzed with Stata 12.0 Software for Windows (Stata Corp, College Station, TX). Infant Hb concentration (g/L) was the main outcome variable in our study. Malaria infection during pregnancy (either placental infection or peripheral parasitaemia) was the main exposure variable. We used the following co-variables in mothers: age (years), ethnic group, socio-economic index, gravidity, gestational age (weeks), number of antenatal visits, BMI, maternal anaemia, iron, folic acid, and vitamin B12 deficiencies, marital status, education level. In the infants, we considered: age (months), sex, low birthweight ((LBW), weight < 2500 g), preterm birth (gestational age < 37 weeks), fever (temperature ≥ 37.5°C), inflammation syndrome, stunting (length-for-age z-score < -2SD), wasting (weight-for-length z-score < -2SD), malaria and helminth infections, sickle cell disease, serum ferritin, folic acid and vitamin B12 concentrations. We first described the general characteristics of the women at delivery and their children at birth, 6, 9 and 12 months. The variations in the distribution of Hb levels between 0, 6, 9 and 12 months were assessed by a Kruskal-Wallis test. Proportions were compared with a Fisher exact test. At each time point (birth, 6, 9 and 12 months), infant Hb levels were compared with Student's t-test or Mann-Withney non-parametric test as appropriate. Then, we assessed the effect of MiP on infant Hb at birth using multivariable linear regression. Afterwards, we used a longitudinal approach to take into account Hb variations over time and to estimate the impact of MiP on infant Hb throughout the first year of life. Assuming that successive Hb measurements in the same infant were correlated, the data presented a hierarchical two-level structure, where Hb measurements (level 1) were clustered within infants (level 2) [20]. Therefore, a linear mixed model with a random intercept was built as specified in the equation below: where haemoglobin (ij) is the ith Hb measurement of infant j, β 00 is the intercept, Xqj is the q explicative variable of infant j and β q0 its associated coefficient, u0j is the random intercept corresponding to the infant-to-infant variation in Hb level [u0j —N (0, π00)], and eij is the residual variation [eij—N(0, σ²)]. Fixed effects parameters were estimated using the maximum likelihood method, and variance components were estimated using the restricted maximum likelihood method. To take into account the numerous physiological changes and the non linear evolution of Hb during the first 3 months of life, we excluded the first Hb measurement (at birth) from the hierarchical mixed model. Also, we adjusted our longitudinal analysis on the hematological parameters of the newborn such as iron or folic acid deficiencies at birth that could modify the levels of Hb during the first year of life. All variables with P values below 0.2 in univariate analysis and which were not in the causal pathway between placental malaria and infant Hb, were included in multivariate analyses. Manual backward selection procedure was performed and statistical significance was set at P < 0.05. For variables with more than two categories, P value of the global test is given. This study was approved by the Ethics Committee of the Health Sciences Faculty of Cotonou in Benin. Before each inclusion, all participants involved in our study provided their written informed consent to participate in this study. The study was also explained in the local language to the participant, and her voluntary consent was obtained. In case the woman could not read, an impartial witness was involved in the process. Mothers were free to interrupt their participation at any time in the study.
