Background
During the past decade, considerable efforts have been made to reduce the global burden of malaria [
1‐
3]. These efforts have led to significant reductions in malaria-related morbidity and mortality [
1‐
5]. Indeed, the World Health Organization (WHO) has estimated that the incidence of malaria has been reduced by 22% over the last 8 years in sub-Saharan Africa, while the figure is close to 70% in South-East Asia [
1].
Although in some continents, several countries have eradicated malaria as endemic infectious disease in the last 50 years, other countries still remain endemic to malaria with varying trends over time, especially in sub-Saharan Africa [
6‐
11]. Indeed, in 2018, 19 sub-Saharan African countries and India bore 85% of the world’s malaria burden, with 94% of malaria-related death occurring in sub-Saharan Africa [
1]. These data point out difficulties in achieving the global technical strategy against malaria 2016–2030, which aims to reduce malaria morbidity by at least 40% in 2020 [
1] This calls for intensified efforts, particularly in the regions of the world most affected by malaria.
Assessment of the level of achievement of malaria control and eradication goals has been possible in recent years thanks to the presence of high-quality global data. Enabling the assessment of the global burden of malaria often at very high spatial resolutions [
2,
4,
5]. Nevertheless, among all published studies reporting on malaria, many focus on adults, pregnant women and children under 5 years of age [
4,
5,
12]. Yet this last age category includes several heterogeneous groups with different immunological characteristics and varying susceptibility to infection [
13‐
15].
While it is recognized that pregnant women and children under 5 years of age are the two categories of the population most vulnerable to malaria, data for children under 5 years of age need to be disaggregated because they include subgroups (neonates, infants, toddlers and early childhood children) that have different immunological characteristics and host-parasite interactions [
13‐
18].
Currently, in countries with high malaria endemicity, the diagnosis of malaria in neonatal period is considered as a differential diagnosis of numerous other infections. It is only mentioned when an infectious work-up has formally ruled out any other aetiology of sepsis in a neonate [
19,
20]. However, given the resurgence of observational studies reporting the presence of asexual forms of
Plasmodium in neonates of mothers infected with malaria on one hand, and of studies reporting malaria as the sole aetiology of neonatal infection in symptomatic infants on the other hand, congenital and neonatal malaria has received renewed interest from healthcare workers in recent years [
20‐
25]. Therefore, to curb the burden of malaria globally with specific attention in endemic areas, it is important to accurately analyse epidemiological data with the aim of informing policy makers, and hence produce tailored public health interventions. The present systematic review with meta-analysis aims to provide a comprehensive overview of clinical malaria occurrence in neonates in endemic settings.
Discussion
This first systematic review with meta-analysis of 28,083 neonates living in 14 endemic malaria countries points out a high prevalence of clinical congenital and neonatal malaria, with a substantial heterogeneity not explained by the geographical location of the study endemic area and diagnostic method used. Considering 1,000 neonates, 40 may experience malaria clinical disease during the first 7 days of their life. After this period, for those aged less than 28 days, among 1000 neonates, 10 may experience malaria clinical disease.
As suggested by several observational studies, the prevalence of neonatal malaria appears to be lower than that observed in adults and children under 5 years of age. Indeed, according to WHO, 11.5% to 46.4% of the cases were thought to occur in the under 5 age group in sub-Saharan Africa [
1]. The relatively low prevalence of clinical malaria in the neonatal period is probably multifactorial [
36,
37]. The presence in the neonatal red blood cell of a high concentration of foetal haemoglobin that induces inadequate intraerythrocytic conditions for
Plasmodium growth is considered one of the most plausible explanations [
37‐
39]. The passively acquired neonatal immunity due to in utero transmission from the mother to the newborn could also play a role in its ability to not develop symptomatic forms of malaria even after exposure to
Plasmodium [
38]. As the kinetics of these passively acquired maternal antibodies show a progressive decline after three to six months, and that the amount of fetal haemoglobin decreases significantly at that age, children older than 6 months are more vulnerable to malaria [
39‐
41]. This may explain why the prevalence of malaria increases in infants after the neonatal period [
40,
41].
Although based on the analysis of two studies done outside Africa against sixteen done on this continent, findings of the current review suggest a lower prevalence of clinical congenital malaria in sub-Saharan Africa compared to regions outside the continent. Given the small number of studies conducted outside Africa, it is necessary to have more studies done on the occurrence of clinical congenital and neonatal malaria in endemic regions outside Africa in order to draw meaningful conclusions. Nevertheless, if these results are confirmed on a larger sample, they could be explained by the fact that in Africa, the prevalence of gestational and placental malaria is higher than elsewhere in the world, as reported by Desai and colleagues, who found that one out of four women in this region presents evidence of placental malaria infection during childbirth [
42]. Similar results were also reported by the WHO in 2019, which estimated that 29% of pregnant women living in areas of high to moderate transmission in sub-Saharan Africa were exposed to malaria [
1]. This triggers the synthesis and transfer of maternal antibodies against
Plasmodium from mother to fetus during pregnancy and confers passive immunity against malaria to the neonates in this area, compared to other regions of the world where pregnant women are less exposed to malaria [
42]. More importantly, a correlation has been shown between maternal and fetal antibody levels [
36]. In addition, some studies have shown a higher prevalence of malaria parasitaemia in neonates in sub-Saharan Africa, compared to those living outside the region [
24,
43‐
48]. They estimated the prevalence of congenital malaria, regardless of the presence of symptoms, to be between 10.8% and 54.2% in sub-Saharan Africa, 12.5% and 18.5% in the province with the highest burden in China [
43,
47,
48]. In the USA, between 1960 and 2005, only 60 cases of congenital malaria were observed, and only a few number of cases were reported in Latin America [
49‐
51]. The clinical picture of neonatal malaria is therefore influenced by the level of endemicity in the country. Children born from mothers living in high endemic areas, and hence passively received antibodies during pregnancy, are less likely to present clinical symptoms of malaria compared to those born from mothers living in non-endemic areas. In the latter, malaria may mimic a neonatal sepsis and represents a life-threatening condition.
Unsurprisingly, most of the studies included in this review are from sub-Saharan Africa, as in many previous reviews on gestational or congenital malaria [
42,
43]. This is mainly due to the fact that 85% of the global burden of malaria is carried by nineteen sub-Saharan African countries and India, [
1] possibly explaining their inclination to produce epidemiological data on clinical malaria in different age categories, including neonates.
In all the studies included in the current review, the diagnosis of malaria was made when the clinical picture was suggestive of infection and Plasmodium was the only infectious agent found in the biological exams. In low-income settings, given the presence of many other parasitic, bacterial and viral life-threatening infections, it is sometimes challenging to incriminate Plasmodium as the main aetiology of an infectious state when other germs are found in the blood especially in neonates. Therefore, to avoid pejorative outcomes, it may be necessary to adjunct anti-malarial therapy to medications meant to cater for other associated infectious agents.
In the current study, malaria prevalence data were pooled regardless of Plasmodium species. This is because only two studies reported the presence of non-Plasmodium falciparum malaria. The low number of studies reporting on non-P. falciparum malaria was probably due to the fact that most of our studies were from sub-Saharan Africa, where P. falciparum is the main aetiology of malaria and is more often associated with clinical manifestations, whereas in America and South-East Asia, Plasmodium vivax and Plasmodium ovale are more prevalent and less likely to produce severe clinical forms compared to P. falciparum.
The results of the current study support the need to consider malaria as a plausible differential diagnosis of fever in neonates, in areas where malaria is still endemic in general, and in sub-Saharan Africa in particular. They also call for the intensification of actions geared towards preventing neonatal malaria by reinforcing therapeutic and preventive measures for pregnant women, and lastly, for the elaboration of specific guidelines for the treatment of malaria in neonates.
Nevertheless, the results of this review must be interpreted in the context of some limitations. Firstly, not all of the included studies reported the proportion of women who experienced malaria during pregnancy and the proportion, who received intermittent preventive treatment. This made it difficult to assess the influence of malaria in pregnancy on the prevalence of clinical congenital and neonatal malaria. Secondly, only three of the studies included in the analysis were conducted outside sub-Saharan Africa, making it difficult to generalize the results to all malaria endemic areas. Thirdly, only two studies reported the presence of P. vivax in their sample, thus a subgroup analysis according to Plasmodium species was not conducted. Nevertheless, this study is the first to comprehensively synthesize epidemiological data on clinical congenital and neonatal malaria in endemic areas. Strong and robust methodological methods were used to minimize bias and produce the most accurate estimates. In addition, no restriction related to language or year of publication were applied, which allowed the analysis of all potentially eligible studies.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.