Exclusive Breastfeeding and Clinical Malaria Risk in 6-Month-Old Infants: A Cross-Sectional Study from Kinshasa, Democratic Republic of the Congo

Introduction

In 2013, more than 80% of the world's malaria cases and malaria-related deaths occurred in the World Health Organization (WHO) African region.1 Most of these deaths occurred in children under the age of 5 years, who lack naturally acquired immunity (NAI) and as a consequence have the highest rates of infection, complications, and mortality.1,2 However, despite lacking NAI, the first months of life are marked by a low incidence of clinical malaria.2,3 The low incidence of clinical malaria among infants has been hypothesized to be due to transplacental transfer of protective maternal immunoglobulin G (IgG),4 although other studies have not supported these findings.3

Similarly, the effect of breastfeeding on malaria risk during the first months of life remains uncertain. Recent studies found increased malaria prevalence among exclusively breastfed infants in Uganda and Malawi, whereas a study conducted in Nigeria found a decreased prevalence.5–7 Despite the different prevalence findings, all three studies suggested that exclusive breastfeeding (EBF) had no significant effect on malaria infection risk.5–7 However, the study from Nigeria was based on questionnaires and is subject to recall bias, whereas the Malawian study did not use the proper EBF definition as proposed by the WHO.7,8 These conflicting results need to be resolved, as the WHO recommends that all mothers EBF for the first 6 months in recognition of EBF benefits for infants.9–11 We sought to determine whether EBF affected the susceptibility to malaria infection and clinical malaria in a cohort of EBF and non-EBF infants residing in Kinshasa, Democratic Republic of the Congo.

Methods

This cross-sectional study was conducted as an ancillary study to a randomized trial designed to evaluate breastfeeding promotion interventions (Yotebieng and others12). The study took place in Kinshasa, the Democratic Republic of the Congo, in six health facilities (Esengo, Kikenda, Kitega, Libikisi, Lukunga, and Luyindu). All mothers who birthed a healthy singleton during the enrollment period (May 24–August 25, 2012) and intended to attend well-baby clinic visits in the same facility were eligible for the original trial. Written consent and dried blood spots (DBS) were collected during the 24-week study visit using Whatman^™ filter paper (GE Healthcare Bio-Sciences, Pittsburgh, PA) and shipped to the University of North Carolina at Chapel Hill for laboratory testing. Of 975 infants, 855 enrolled in the parent study were seen at the 24-week visit. However, due to limited supplies, DBS samples were collected for only 495 infants. The human immunodeficiency virus (HIV) status for 480/495 mothers was obtained from the Prevention of Mother-to-Child Transmission registry.

From each patient's DBS card, genomic DNA (gDNA) was extracted as previously described.13 The samples were then evaluated using real-time polymerase chain reaction (PCR) with each well containing 2 μL of extracted gDNA, 6.25 μL of FastStart Universal Probe Master (Rox; Roche Diagnostics, Indianapolis, IN), 155 nL of 20 μM forward and reverse Plasmodium falciparum lactate dehydrogenase (PfLDH) primer, 190 nL of 20 μM PfLDH TaqMan^® Probe (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA), and 3.25 μL of molecular-grade water. All plates included a negative control and five positive controls (P. falciparum DNA) at sequential log concentrations ranging from 10⁻² to 10⁻⁵ ng/μL.

Samples were considered positive if they had a C_t of < 38 or two C_t values of < 39 using the Bio-Rad CFX384^™ Real-Time PCR Detection system (Bio-Rad, Clinical Diagnostics, Hercules, CA). Positive samples in the first series or indeterminates were evaluated in duplicate or, if necessary, triplicate using increased concentrations of gDNA (4 μL).

Information on infant feeding was collected through a face-to-face interview during the 24-week visit. Infants were classified as being exclusively breastfed using the definitions proposed by the WHO14 and recall of the past 24 hours and past 7-day time frames. In addition, mothers were asked during the 24-week interview whether their infant had experienced any fever or had consumed any medications during the month preceding the interview. Mothers were also asked if their infant had slept under a bed net the night preceding the interview. Malaria microscopy was not performed during the 24-week clinical visit and infant's temperatures were not recorded.

The two outcomes of interest for the analysis were malaria infection, which was defined by a positive PCR (since microscopy was not performed), and clinical malaria, which was defined as a positive PCR and reported fever. Other variables considered in the analysis were related to the mother's sociodemographic characteristics (Table 1).

Associations between potential confounders and the outcomes were assessed using Fisher's exact test or χ² test, when appropriate, for categorical variables and Wilcoxon rank sum test for continuous variables. Simple and stratified exact Cochran–Mantel–Haenszel odds ratios (ORs) and 95% confidence intervals (95% CIs) were used to assess the strength of the association between EBF and the two outcomes of interest. Analyses were performed using SAS^®9.3 (Cary, NC) and assessed at a 0.05 significance level.

Results

At the 24-week interview time point, 137 (27.7%) of the infants were EBF. First-time mothers (primiparous) and younger mothers (median = 27; interquartile range = 22.0–32.0) were less likely to exclusively breastfeed (P = 0.0078 and P = 0.0141, respectively). In addition, EBF infants were more likely to have used bed nets (P < 0.0001) but less likely to have used antimalarials (P = 0.0439).

Malaria parasitemia was confirmed by PCR in 52 (10.5%) infants, of whom, 21 were reported to have had a fever in the last month. In bivariate analyses, EBF was associated with a slight, but not significant, reduced risk of malaria infection (prevalence of 9.5% versus 10.9%; OR = 0.82; 95% CI = 0.39–1.62). EBF significantly reduced the risk of clinical malaria (OR = 0.13; 95% CI = 0.00–0.80). Indeed, only one (0.7%) case of clinical malaria was in an EBF infant. Bed net use was associated with reduced risk of malaria infection (OR = 0.63; 95% CI = 0.34–1.19) and clinical malaria (OR = 0.41; 95% CI = 0.15–1.09). The reported use of antimalarial drugs was associated with increased risk of malaria infection (OR = 1.76; 95% CI = 0.71–3.97) and clinical malaria (OR = 6.81; 95% CI = 2.39–18.57). Maternal parity appeared to have little, if any, effect on the rate of malaria infection or clinical malaria (Table 1).

Among the infants without malaria infections, 116 (26.19%) presented with fever. EBF was found to be protective against fever among infants with and without malaria infection (OR = 0.08; 95% CI = 0.01–0.67 and OR = 0.74; 95% CI = 0.432–1.23, respectively).

In stratified analyses, among children who used bed nets, 10 (9.3%) EBF and eight (8.7%) non-EBF infants presented with malaria infection (OR = 1.07; 95% CI = 0.42–2.56; Table 2). Among non-bed net users, malaria infection was found in three EBF (10.3%) and 21 non-EBF (13.9%) infants (OR = 0.71; 95% CI = 0.12–2.67). Overall, controlling for bed net use, there was a tendency toward lower odds of malaria infection among EBF infants (OR = 0.86; 95% CI = 0.41–1.71), although these associations lacked statistical significance. For clinical malaria, among bed net users, one (0.9%) EBF and eight (3.9%) non-EBF infants presented with clinical malaria (OR = 0.23; 95% CI = 0.01–1.78). Among non-bed net users, all 12 cases of clinical malaria recorded were among non-EBF infants. Stratification by bed net use showed decreased odds of clinical malaria among EBF infants (OR = 0.12; 95% CI = 0.00–0.79; Table 2). Similar results were obtained after stratification by use of antimalarial drugs (Table 2). Of 480 mothers, four with known HIV statuses were HIV positive and only one infant with an HIV-positive mother demonstrated malaria infection (zero cases of clinical malaria).

Discussion

The overall prevalence of malaria infection among the 495 6-month-old infants from Kinshasa, Democratic Republic of the Congo, considered in this study was 10.5%. A concordance of point estimates stratified by bed net use, reported fever, and other factors suggested that EBF at 24 weeks was associated with lower odds of developing clinical malaria, although only the findings of EBF protecting against fever among infants with malaria infection and EBF reducing the odds of clinical malaria were statistically significant.

This reduction in malaria susceptibility among EBF infants may be related to the immunomodulatory effects of breast milk.15 Previous studies that found a lack of an association between EBF and malaria did not use PCR for malaria detection, measured malaria infection in a broad interval (age 3–12 months), and did not consider clinical malaria as an outcome.

This cross-sectional study has several limitations, including the small number of malaria cases, the unknown status of genetic factors that affect disease susceptibility (i.e., hemoglobinopathies), and the potential of recall bias. The issue of recall bias and misclassification was attenuated in that mothers and interviewers were blinded to the PCR results during the interview. As a result, any misclassification would be nondifferential and bias the association toward the null. In addition, gravidity was not associated with infection prevalence (Table 1), which suggests primigravida mothers were not subject to increased recall bias. This lack of association between parity and malaria infection prevalence may have resulted from the disproportionate number of primigravida mothers in the study (82.6%). This high number of primigravidae women in our study may limit its generalizability.

Although it is unlikely that maternal breast milk IgG antibodies are systemically absorbed into the circulation where malaria parasites reside,16 EBF protected infants in this study from developing clinical malaria. This observed effect may be a byproduct of EBF reducing overall disease burden,10,11 allowing for increased resource allocation to malaria defenses, or the role of EBF in training and modulating the infant's immune system.15 It is possible that some of the fever episodes among the malaria-infected children were due to other etiologic agents. As such, our definition of clinical malaria is lenient. In addition, antimalarial use was associated with increased risk of malaria infection, possibly due to detection of biologically inactive parasites. However, the point estimates of both malaria infection and clinical malaria in the presence and absence of bed nets suggest that EBF has a protective effect. In addition, EBF reduced fever risk among infants with malaria infection more than those without a malaria infection (OR = 0.08 versus OR = 0.74, respectively). Given these indications, our findings need to be replicated in a larger prospective study.